diff --git a/.github/workflows/realm-empty-env.yml b/.github/workflows/realm-empty-env.yml deleted file mode 100644 index 9150fee5a0..0000000000 --- a/.github/workflows/realm-empty-env.yml +++ /dev/null @@ -1,25 +0,0 @@ -name: Realm Deploy to Empty Environment - -on: - push: - branches: - - staging - workflow_dispatch: - -jobs: - push: - runs-on: ubuntu-latest - environment: empty - steps: - - uses: actions/checkout@v2 - - name: "Install the Realm CLI" - run: | - npm install -g mongodb-realm-cli - - name: Login - run: | - realm-cli login --api-key="${{ secrets.REALM_API_PUBLIC_KEY }}" --private-api-key="${{ secrets.REALM_API_PRIVATE_KEY }}" --realm-url https://services.cloud.mongodb.com --atlas-url https://cloud.mongodb.com - - name: Push - run: | - cd site/realm - realm-cli push --remote="${{ secrets.GATSBY_REALM_APP_ID }}" -y - diff --git a/site/gatsby-site/gatsby-config.js b/site/gatsby-site/gatsby-config.js index d3ac502d00..7d1f5ec68c 100755 --- a/site/gatsby-site/gatsby-config.js +++ b/site/gatsby-site/gatsby-config.js @@ -11,6 +11,12 @@ cloudinary.config({ cloud_name: config.cloudinary.cloudName }); const adapter = require('gatsby-adapter-netlify').default; +let googleTrackingIds = []; + +if (process.env.SITE_URL === config.gatsby.siteUrl) { + googleTrackingIds.push(config.gatsby.gaTrackingId); +} + const plugins = [ 'layout', { @@ -65,7 +71,7 @@ const plugins = [ { resolve: `gatsby-plugin-google-gtag`, options: { - trackingIds: [config.gatsby.gaTrackingId], + trackingIds: googleTrackingIds, }, }, { diff --git a/site/gatsby-site/i18n/locales/es/entities.json b/site/gatsby-site/i18n/locales/es/entities.json index bb668d7b61..ad0e9a5f5e 100644 --- a/site/gatsby-site/i18n/locales/es/entities.json +++ b/site/gatsby-site/i18n/locales/es/entities.json @@ -21,8 +21,8 @@ "Incidents involved as both Developer and Deployer": "Incidentes involucrados como desarrollador e implementador", "Incidents Harmed By": "Afectado por Incidentes", "Alleged: <2> developed and deployed an AI system, which harmed <6>.": "Presunto: un sistema de IA desarrollado e implementado por <2>, perjudicó a <6>.", - "The AI implicated system is": "El sistema de IA implicado es", - "The AI implicated systems are": "Los sistemas de IA implicados son", + "Implicated AI system:": "Sistema de IA implicado:", + "Implicated AI systems:": "Sistemas de IA implicados:", "Alleged: <1> developed an AI system deployed by <4>, which harmed <6>.": "Presunto: un sistema de IA desarrollado por <1> e implementado por <4>, perjudicó a <6>.", "Entities involved in AI Incidents": "^Entities involved in AI Incidents", "{{count}} Incident responses": "{{count}} respuestas de incidentes", diff --git a/site/gatsby-site/i18n/locales/fr/entities.json b/site/gatsby-site/i18n/locales/fr/entities.json index d780b45678..6eac29c7fc 100644 --- a/site/gatsby-site/i18n/locales/fr/entities.json +++ b/site/gatsby-site/i18n/locales/fr/entities.json @@ -28,6 +28,6 @@ "Back to Entity: {{name}}": "Retour à l'entité: {{name}}", "Entity updated successfully.": "Entité mise à jour avec succès.", "Error updating Entity.": "Erreur lors de la mise à jour de l'entité.", - "The AI implicated system is": "Le système d'IA impliqué est", - "The AI implicated systems are": "Les systèmes d'IA impliqués sont" + "Implicated AI system:": "Système de l'IA impliqué:", + "Implicated AI systems:": "Systèmes de l'IA impliqués:" } diff --git a/site/gatsby-site/i18n/locales/ja/entities.json b/site/gatsby-site/i18n/locales/ja/entities.json index 2019b640ea..462109f69a 100644 --- a/site/gatsby-site/i18n/locales/ja/entities.json +++ b/site/gatsby-site/i18n/locales/ja/entities.json @@ -29,6 +29,6 @@ "Back to Entity: {{name}}": "組織に戻る: {{name}}", "Entity updated successfully.": "組織が正常に更新されました。", "Error updating Entity.": "組織の更新中にエラーが発生しました。", - "The AI implicated system is": "関連するAIシステムは です", - "The AI implicated systems are": "関連するAIシステムは です" + "Implicated AI system:": "AI関連システム:", + "Implicated AI systems:": "AI関連システム:" } diff --git a/site/gatsby-site/playwright/config.ts b/site/gatsby-site/playwright/config.ts index a5070eec45..c6fc207f14 100644 --- a/site/gatsby-site/playwright/config.ts +++ b/site/gatsby-site/playwright/config.ts @@ -5,6 +5,7 @@ type ConfigType = { AVAILABLE_LANGUAGES?: string; NEXTAUTH_URL?: string; NEXTAUTH_SECRET?: string; + SITE_URL?: string; }; const config: ConfigType = { @@ -14,6 +15,7 @@ const config: ConfigType = { AVAILABLE_LANGUAGES: process.env.GATSBY_AVAILABLE_LANGUAGES ?? '', NEXTAUTH_URL: process.env.NEXTAUTH_URL!, NEXTAUTH_SECRET: process.env.NEXTAUTH_SECRET!, + SITE_URL: process.env.SITE_URL ?? '', } Object.keys(config).forEach((key) => { diff --git a/site/gatsby-site/playwright/e2e-full/cite.spec.ts b/site/gatsby-site/playwright/e2e-full/cite.spec.ts index 3fe902cb9a..103aa882fb 100644 --- a/site/gatsby-site/playwright/e2e-full/cite.spec.ts +++ b/site/gatsby-site/playwright/e2e-full/cite.spec.ts @@ -460,7 +460,7 @@ test.describe('Cite pages', () => { test('Should show proper entities card text', async ({ page }) => { await page.goto('/cite/3/'); await expect(page.locator('[data-cy="alleged-entities"]')).toHaveText( - 'Alleged: Kronos developed an AI system deployed by Starbucks, which harmed Starbucks Employees.The AI implicated system is Entity 1.' + 'Alleged: Kronos developed an AI system deployed by Starbucks, which harmed Starbucks Employees.Implicated AI system: Entity 1' ); }); diff --git a/site/gatsby-site/playwright/e2e-full/dynamicCite.spec.ts b/site/gatsby-site/playwright/e2e-full/dynamicCite.spec.ts index 28801c3ce3..1905f2f9d6 100644 --- a/site/gatsby-site/playwright/e2e-full/dynamicCite.spec.ts +++ b/site/gatsby-site/playwright/e2e-full/dynamicCite.spec.ts @@ -34,7 +34,7 @@ test.describe('Dynamic Cite pages', () => { await expect(page.getByText(`Kronos’s scheduling algorithm and its use by Starbucks managers allegedly negatively impacted financial and scheduling stability for Starbucks employees, which disadvantaged wage workers.`)).toBeVisible(); await expect(page.locator('[data-cy="alleged-entities"]')).toHaveText( - 'Alleged: Kronos developed an AI system deployed by Starbucks, which harmed Starbucks Employees.The AI implicated system is Entity 1.' + 'Alleged: Kronos developed an AI system deployed by Starbucks, which harmed Starbucks Employees.Implicated AI system: Entity 1' ); await expect(page.locator('[data-cy="citation"]').getByText("Report Count", { exact: true }).locator('xpath=following-sibling::div[1]')).toHaveText('2'); diff --git a/site/gatsby-site/playwright/e2e-full/translationBadge.spec.ts b/site/gatsby-site/playwright/e2e-full/translationBadge.spec.ts index c86902cf79..b6fe0269f0 100644 --- a/site/gatsby-site/playwright/e2e-full/translationBadge.spec.ts +++ b/site/gatsby-site/playwright/e2e-full/translationBadge.spec.ts @@ -1,5 +1,5 @@ import { expect } from '@playwright/test'; -import { test } from '../utils'; +import { mockAlgolia, test } from '../utils'; test.describe('Translation Badges', () => { test('Should be visible on blog post', async ({ page }) => { @@ -15,9 +15,18 @@ test.describe('Translation Badges', () => { await expect(page.locator('a', { hasText: 'Ver Original' })).not.toBeVisible(); }); - test('Should be visible on the discover app', async ({ page, skipOnEmptyEnvironment }) => { - await page.goto('/es/apps/discover?display=details&incident_id=1&page=1&source_domain=today.com'); - await expect(page.locator('[data-cy="5d34b8c29ced494f010ed45c"]').locator('[data-cy="translation-badge"]').getByText('Traducido por IA')).toBeVisible(); + test('Should be visible on the discover app if the item has a translation available', async ({ page, skipOnEmptyEnvironment }) => { + await mockAlgolia(page); + + await page.goto('/es/apps/discover'); + await expect(page.locator('[data-cy="6243a9eedf8b4b62d982817e"]').locator('[data-cy="translation-badge"]').getByText('Traducido por IA')).toBeVisible(); + }); + + test('Should not be visible on the discover app if the item does not have a translation available', async ({ page, skipOnEmptyEnvironment }) => { + await mockAlgolia(page); + + await page.goto('/es/apps/discover'); + await expect(page.locator('[data-cy="5d34b8c29ced494f010ed470"]').locator('[data-cy="translation-badge"]').getByText('Traducido por IA')).not.toBeVisible(); }); test('Should be visible on an report card on the citation page if it was translated', async ({ page, skipOnEmptyEnvironment }) => { diff --git a/site/gatsby-site/playwright/e2e/unit/gtag.spec.ts b/site/gatsby-site/playwright/e2e/unit/gtag.spec.ts new file mode 100644 index 0000000000..9ec3b96187 --- /dev/null +++ b/site/gatsby-site/playwright/e2e/unit/gtag.spec.ts @@ -0,0 +1,40 @@ +import { expect } from '@playwright/test'; +import config from '../../../config'; +import { test } from '../../utils'; + +const googleTrackingId = config.gatsby.gaTrackingId; + +test.describe('Google Analytics Tracking', () => { + test('Should track gtag only on production', async ({ page, runOnlyInProduction }) => { + await page.goto('/'); + + const hasGoogleTracking = await page.locator(`script[src*="https://www.googletagmanager.com/gtag/js?id=${googleTrackingId}"]`).count(); + + expect(hasGoogleTracking).toBeGreaterThan(0); + + const [trackingRequest] = await Promise.all([ + page.waitForRequest((request) => request.url().includes(`https://www.google-analytics.com`) && request.method() === 'POST'), + + ]); + + expect(trackingRequest).toBeTruthy(); + }); + + + test('Should not track gtag outside production', async ({ page, runAnywhereExceptProduction }) => { + await page.goto('/'); + + const hasGoogleTracking = await page.locator(`script[src*="https://www.googletagmanager.com/gtag/js?id=${googleTrackingId}"]`).count(); + expect(hasGoogleTracking).toBe(0); + + let trackingRequestMade = false; + page.on('request', (request) => { + if (request.url().includes(`https://www.google-analytics.com`) && request.method() === 'POST') { + trackingRequestMade = true; + } + }); + + await page.reload(); + expect(trackingRequestMade).toBe(false); + }); +}); diff --git a/site/gatsby-site/playwright/fixtures/algoliaMock.ts b/site/gatsby-site/playwright/fixtures/algoliaMock.ts index 645360afce..02fafd361c 100644 --- a/site/gatsby-site/playwright/fixtures/algoliaMock.ts +++ b/site/gatsby-site/playwright/fixtures/algoliaMock.ts @@ -1,7772 +1,7709 @@ export const algoliaMock = { - results: [ + results: [ + { + hits: [ { - "hits": [ - { - "authors": [ - "Aimee Picchi" - ], - "description": "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", - "epoch_date_downloaded": 1555113600, - "epoch_date_modified": 1725431608, - "epoch_date_published": 1443052800, - "epoch_date_submitted": 1559347200, - "image_url": "https://cbsnews1.cbsistatic.com/hub/i/r/2015/03/17/01a38576-5108-40f7-8df8-5416164ed878/thumbnail/1200x630/ca8d35fe6bc065b5c9a747d92bc6d94c/154211248.jpg", - "language": "en", - "report_number": 23, - "source_domain": "cbsnews.com", - "submitters": [ - "Catherine Olsson" - ], - "title": "​Is Starbucks shortchanging its baristas?", - "name": "​Is Starbucks shortchanging its baristas?", - "url": "https://www.cbsnews.com/news/is-starbucks-shortchanging-its-baristas/", - "tags": [], - "editor_notes": "", - "cloudinary_id": "reports/cbsnews1.cbsistatic.com/hub/i/r/2015/03/17/01a38576-5108-40f7-8df8-5416164ed878/thumbnail/1200x630/ca8d35fe6bc065b5c9a747d92bc6d94c/154211248.jpg", - "text": "For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health insurance and some extra money while he studies at Georgia Perimeter College. What it doesn't provide is the kind of stable schedule that the company promised its workers last year.\n\n\"It's the wild inconsistency\" of the hours that's a problem, Weisse, 32, said. \"We're supposed to get them 10 days in advance, which often happens, but there's no guarantee. If our manager doesn't get it to us on time, we just have to deal with it.\"\n\nThat became a problem recently when Weisse's manager gave him only a few days notice on his work hours, which ended up conflicting with an anatomy and physiology exam at his college. Weisse ended up paying another worker $20 to take his shift so he could take the exam.\n\nThe short notice is especially frustrating because of Starbucks' vow last year to post employees' schedules at least 10 days in advance, as well as the company's insistence that workers provide at least one-month notice when they need to take a day off.\n\nWhat's behind Starbucks price increases?\n\nWeisse isn't alone in complaining that Starbucks isn't living up to its promises to overhaul its labor practices for its roughly 130,000 baristas. That vow followed an article last year by The New York Times that detailed how workers were struggling to manage childcare and other obligations when the company provided only a few days notice about their schedules.\n\nAbout half of roughly 200 Starbucks baristas in a recent survey said they are still receiving their schedule with less than one week's notice. Others also reported being asked to handle \"clopens,\" split shifts in which employees work a closing shift late into the evening and then an early opening shift the following morning. The company last year promised to end the practice.\n\nOf course, Starbucks isn't alone in using \"just-in-time\" scheduling, with the retail and restaurant industry increasingly turning to software that allows them to change work schedules at the last minute, depending on whether business picks up or slows down. But it is Starbucks that has become a lightning rod on the issue given its vows to improve how it treats employees and its own emphatic claims to valuing workers, whom it labels \"partners.\"\n\n\"Starbucks has the values and wants to do right by their employees,\" said Carrie Gleason, director of the Fair Workweek Initiative at the Center for Popular Democracy, an advocacy group focused on workers'rights, and a co-author of the group's new report on the company's labor practices. \"However, since last year when the company recognized there was a serious problem with the way it scheduled workers and pledged to reform, still so many of the same issues persist.\"\n\nStarbucks didn't respond to requests for comment on the study or on baristas' reports of labor practices that are failing to meet the company's stated goals.\n\nIn an internal memo this week published by Time, Starbucks executive Cliff Burrows wrote that the company couldn't validate the survey, but added that \"the findings suggest, contrary to the expectations we have in place, that some partners are receiving their schedules less than one week in advance and that there is a continuing issue with some partners working a close and then an opening shift the following morning.\" He asks store managers \"to go the extra mile to ensure partners have a consistent schedule.\"\n\nStarbucks ends \"race together\" campaign amid public backlash\n\nTo be sure, some Starbucks workers are receiving at least 10 days notice on their work hours, with the survey finding that about one-third receive two weeks notice and another 18 percent get their schedules three weeks in advance. But that leaves almost half of workers who only receive one week's notice, making it more difficult from them manage other obligations, such as school, family commitments or other jobs.\n\nClopens remain a problem, as well. About 60 percent of workers who have to handle a clopen receive seven or fewer hours of rest between a closing and an opening shift, the study found.\n\nThat's prompted one former Starbucks employee to start a petition to end the practice of scheduling clopens. Ciara Moran noted in her petition that she sometimes was only able to get four or five hours of sleep on the days she was scheduled for clopens. She said she quit her job because she doubted whether it was possible to get ahead given the demands on workers.\n\nEven if Starbucks stuck with its policy of providing eight hours between shifts, that's not enough time, especially given that many workers in the service sector have long commutes, the study said.\n\nAnother issue singled out by the report is Starbucks' practices on sick time. Since paid time off is only available to workers with at least a year on the job, about 40 percent of employees in the survey said they had dealt with barriers in taking sick days.\n\nIn a perfect world, Weisse said he'd like to receive his schedule either a month or a", - "mongodb_id": "5d34b8c29ced494f010ed470", - "featured": 2, - "flag": true, - "is_incident_report": true, - "namespaces": [ - "CSETv0", - "GMF", - "CSETv1_Annotator-1", - "CSETv1" - ], - "classifications": [ - "CSETv0:Location:Global", - "CSETv0:Near Miss:Unclear/unknown", - "CSETv0:Named Entities:Starbucks", - "CSETv0:Named Entities:Kronos", - "CSETv0:Technology Purveyor:Starbucks", - "CSETv0:Intent:Unclear", - "CSETv0:Severity:Negligible", - "CSETv0:Harm Type:Psychological harm", - "CSETv0:Infrastructure Sectors:Food and agriculture", - "GMF:Known AI Technical Failure:Underspecification", - "GMF:Potential AI Technical Failure:Tuning Issues", - "GMF:Potential AI Technical Failure:Misconfigured Aggregation", - "GMF:Potential AI Technology:Regression", - "GMF:Potential AI Technology:Diverse Data", - "GMF:Known AI Goal:Market Forecasting", - "GMF:Known AI Goal:Scheduling", - "CSETv1_Annotator-1:Physical Objects:no", - "CSETv1_Annotator-1:Entertainment Industry:no", - "CSETv1_Annotator-1:Report, Test, or Study of data:no", - "CSETv1_Annotator-1:Deployed:yes", - "CSETv1_Annotator-1:Producer Test in Controlled Conditions:no", - "CSETv1_Annotator-1:Producer Test in Operational Conditions:no", - "CSETv1_Annotator-1:User Test in Controlled Conditions:no", - "CSETv1_Annotator-1:User Test in Operational Conditions:no", - "CSETv1_Annotator-1:Tangible Harm:tangible harm definitively occurred", - "CSETv1_Annotator-1:AI System:yes", - "CSETv1_Annotator-1:AI Harm Level:AI tangible harm event", - "CSETv1_Annotator-1:Impact on Critical Services:no", - "CSETv1_Annotator-1:Rights Violation:no", - "CSETv1_Annotator-1:Involving Minor:no", - "CSETv1_Annotator-1:Detrimental Content:no", - "CSETv1_Annotator-1:Protected Characteristic:no", - "CSETv1_Annotator-1:Harm Distribution Basis:none", - "CSETv1_Annotator-1:Clear link to Technology:yes", - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no", - "CSETv1_Annotator-1:Sector of Deployment:wholesale and retail trade", - "CSETv1_Annotator-1:Sector of Deployment:accommodation and food service activities", - "CSETv1_Annotator-1:Public Sector Deployment:no", - "CSETv1_Annotator-1:Autonomy Level:Autonomy2", - "CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1_Annotator-1:Special Interest Intangible Harm:no", - "CSETv1_Annotator-1:Date of Incident Year:2014", - "CSETv1_Annotator-1:Multiple AI Interaction:no", - "CSETv1_Annotator-1:Embedded:no", - "CSETv1_Annotator-1:Location Country (two letters):US", - "CSETv1_Annotator-1:Location Region:North America", - "CSETv1_Annotator-1:Data Inputs:schedules", - "CSETv1_Annotator-1:Data Inputs:worker profiles", - "CSETv1_Annotator-1:AI Task:scheduling", - "CSETv1_Annotator-1:AI Task:productivity optimization", - "CSETv1:Harm Distribution Basis:none", - "CSETv1:Sector of Deployment:accommodation and food service activities", - "CSETv1:Physical Objects:no", - "CSETv1:Entertainment Industry:no", - "CSETv1:Report, Test, or Study of data:no", - "CSETv1:Deployed:yes", - "CSETv1:Producer Test in Controlled Conditions:no", - "CSETv1:Producer Test in Operational Conditions:no", - "CSETv1:User Test in Controlled Conditions:no", - "CSETv1:User Test in Operational Conditions:no", - "CSETv1:Tangible Harm:tangible harm definitively occurred", - "CSETv1:AI System:yes", - "CSETv1:AI Harm Level:AI tangible harm event", - "CSETv1:Impact on Critical Services:no", - "CSETv1:Rights Violation:no", - "CSETv1:Involving Minor:no", - "CSETv1:Detrimental Content:no", - "CSETv1:Protected Characteristic:no", - "CSETv1:Clear link to Technology:yes", - "CSETv1:Annotator’s AI special interest intangible harm assessment:no", - "CSETv1:Public Sector Deployment:no", - "CSETv1:Autonomy Level:Autonomy2", - "CSETv1:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1:Special Interest Intangible Harm:no", - "CSETv1:Date of Incident Year:2014", - "CSETv1:Multiple AI Interaction:no", - "CSETv1:Embedded:no", - "CSETv1:Location Country (two letters):US", - "CSETv1:Location Region:North America", - "CSETv1:Data Inputs:schedules", - "CSETv1:Data Inputs:worker profiles", - "CSETv1:Data Inputs:store traffic", - "CSETv1:AI Task:scheduling", - "CSETv1:AI Task:productivity optimization", - "CSETv1:AI Task:predict store traffic" - ], - "CSETv0": { - "Location": "Global", - "Near Miss": "Unclear/unknown", - "Named Entities": [ - "Starbucks", - "Kronos" - ], - "Technology Purveyor": [ - "Starbucks" - ], - "Intent": "Unclear", - "Severity": "Negligible", - "Harm Type": [ - "Psychological harm" - ], - "Lives Lost": false, - "Harm Distribution Basis": [], - "Infrastructure Sectors": [ - "Food and agriculture" - ], - "Financial Cost": "", - "System Developer": [], - "Sector of Deployment": [ - "" - ], - "Public Sector Deployment": false, - "Nature of End User": "", - "Level of Autonomy": "", - "Relevant AI functions": [], - "AI Techniques": [], - "AI Applications": [], - "Physical System": [], - "Problem Nature": [] - }, - "GMF": { - "Known AI Technical Failure": [ - "Underspecification" - ], - "Known AI Technical Failure Snippets": [ - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"“You’re waiting on your job to control your life,” she said, with the scheduling software used by her employer dictating everything from “how much sleep Gavin will get to what groceries I’ll be able to buy this month.”\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Underspecification\"]" - } - ] - }, - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"Along with virtually every major retail and restaurant chain, Starbucks relies on software that choreographs workers in precise, intricate ballets, using sales patterns and other data to determine which of its 130,000 baristas are needed in its thousands of locations and exactly when.\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Underspecification\"]" - } - ] - }, - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"Among other changes, the company said it would end the practice of \\\"clopening,\\\" when an employee responsible for closing a store late at night is also assigned to open it early in the morning.\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Underspecification\"]" - } - ] - }, - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"In a follow-up piece, the author, Jodi Kantor, points directly to Kronos' scheduling software as the root of the problem.\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Underspecification\"]" - } - ] - } - ], - "Potential AI Technical Failure": [ - "Tuning Issues", - "Misconfigured Aggregation" - ], - "Potential AI Technical Failure Snippets": [ - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"In addition, Kronos is improving a feature meant to help give employees more control over their schedules: Though the software already incorporates employee availability and preferences into its scheduling calculations, improvements to a shift-swapping feature on its employee-facing web and mobile apps will theoretically allow employees to work around conflicts among themselves.\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Tuning Issues\",\"Misconfigured Aggregation\"]" - } - ] - } - ], - "Potential AI Technology": [ - "Regression", - "Diverse Data" - ], - "Potential AI Technology Snippets": [ - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"“You’re waiting on your job to control your life,” she said, with the scheduling software used by her employer dictating everything from “how much sleep Gavin will get to what groceries I’ll be able to buy this month.”\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Regression\"]" - } - ] - }, - { - "attributes": [ - { - "short_name": "Snippet Text", - "value_json": "\"In a follow-up piece, the author, Jodi Kantor, points directly to Kronos' scheduling software as the root of the problem.\"" - }, - { - "short_name": "Related Classifications", - "value_json": "[\"Diverse Data\"]" - } - ] - } - ], - "Known AI Goal": [ - "Market Forecasting", - "Scheduling" - ], - "Known AI Goal Snippets": [] - }, - "CSETv1_Annotator-1": { - "Physical Objects": "no", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "no", - "Deployed": "yes", - "Producer Test in Controlled Conditions": "no", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "no", - "Tangible Harm": "tangible harm definitively occurred", - "AI System": "yes", - "AI Harm Level": "AI tangible harm event", - "Impact on Critical Services": "no", - "Rights Violation": "no", - "Involving Minor": "no", - "Detrimental Content": "no", - "Protected Characteristic": "no", - "Harm Distribution Basis": [ - "none" - ], - "Clear link to Technology": "yes", - "Annotator’s AI special interest intangible harm assessment": "no", - "Sector of Deployment": [ - "wholesale and retail trade", - "accommodation and food service activities" - ], - "Public Sector Deployment": "no", - "Autonomy Level": "Autonomy2", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI tools and methods": "", - "Special Interest Intangible Harm": "no", - "Date of Incident Year": "2014", - "Multiple AI Interaction": "no", - "Embedded": "no", - "Location Country (two letters)": "US", - "Location Region": "North America", - "Infrastructure Sectors": [], - "Operating Conditions": "", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "schedules", - "worker profiles" - ], - "Physical System Type": "", - "AI Task": [ - "scheduling", - "productivity optimization" - ] - }, - "CSETv1": { - "Harm Distribution Basis": [ - "none" - ], - "Sector of Deployment": [ - "accommodation and food service activities" - ], - "Physical Objects": "no", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "no", - "Deployed": "yes", - "Producer Test in Controlled Conditions": "no", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "no", - "Tangible Harm": "tangible harm definitively occurred", - "AI System": "yes", - "AI Harm Level": "AI tangible harm event", - "Impact on Critical Services": "no", - "Rights Violation": "no", - "Involving Minor": "no", - "Detrimental Content": "no", - "Protected Characteristic": "no", - "Clear link to Technology": "yes", - "Annotator’s AI special interest intangible harm assessment": "no", - "Public Sector Deployment": "no", - "Autonomy Level": "Autonomy2", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI tools and methods": "", - "Special Interest Intangible Harm": "no", - "Date of Incident Year": "2014", - "Multiple AI Interaction": "no", - "Embedded": "no", - "Location Country (two letters)": "US", - "Location Region": "North America", - "Infrastructure Sectors": [], - "Operating Conditions": "", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "schedules", - "worker profiles", - "store traffic" - ], - "Physical System Type": "", - "AI Task": [ - "scheduling", - "productivity optimization", - "predict store traffic" - ] - }, - "incident_id": 1, - "incident_date": "2014-08-14", - "epoch_incident_date": 1407974400, - "incident_title": "Kronos Scheduling Algorithm Allegedly Caused Financial Issues for Starbucks Employees", - "incident_description": "Kronos’s scheduling algorithm and its use by Starbucks managers allegedly negatively impacted financial and scheduling stability for Starbucks employees, which disadvantaged wage workers.", - "objectID": "23", - "_snippetResult": { - "description": { - "value": "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", - "matchLevel": "none" - }, - "text": { - "value": "For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health", - "matchLevel": "none" - } - }, - "_highlightResult": { - "description": { - "value": "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", - "matchLevel": "none", - "matchedWords": [] - }, - "title": { - "value": "​Is Starbucks shortchanging its baristas?", - "matchLevel": "none", - "matchedWords": [] - }, - "text": { - "value": "For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health insurance and some extra money while he studies at Georgia Perimeter College. What it doesn't provide is the kind of stable schedule that the company promised its workers last year.\n\n\"It's the wild inconsistency\" of the hours that's a problem, Weisse, 32, said. \"We're supposed to get them 10 days in advance, which often happens, but there's no guarantee. If our manager doesn't get it to us on time, we just have to deal with it.\"\n\nThat became a problem recently when Weisse's manager gave him only a few days notice on his work hours, which ended up conflicting with an anatomy and physiology exam at his college. Weisse ended up paying another worker $20 to take his shift so he could take the exam.\n\nThe short notice is especially frustrating because of Starbucks' vow last year to post employees' schedules at least 10 days in advance, as well as the company's insistence that workers provide at least one-month notice when they need to take a day off.\n\nWhat's behind Starbucks price increases?\n\nWeisse isn't alone in complaining that Starbucks isn't living up to its promises to overhaul its labor practices for its roughly 130,000 baristas. That vow followed an article last year by The New York Times that detailed how workers were struggling to manage childcare and other obligations when the company provided only a few days notice about their schedules.\n\nAbout half of roughly 200 Starbucks baristas in a recent survey said they are still receiving their schedule with less than one week's notice. Others also reported being asked to handle \"clopens,\" split shifts in which employees work a closing shift late into the evening and then an early opening shift the following morning. The company last year promised to end the practice.\n\nOf course, Starbucks isn't alone in using \"just-in-time\" scheduling, with the retail and restaurant industry increasingly turning to software that allows them to change work schedules at the last minute, depending on whether business picks up or slows down. But it is Starbucks that has become a lightning rod on the issue given its vows to improve how it treats employees and its own emphatic claims to valuing workers, whom it labels \"partners.\"\n\n\"Starbucks has the values and wants to do right by their employees,\" said Carrie Gleason, director of the Fair Workweek Initiative at the Center for Popular Democracy, an advocacy group focused on workers'rights, and a co-author of the group's new report on the company's labor practices. \"However, since last year when the company recognized there was a serious problem with the way it scheduled workers and pledged to reform, still so many of the same issues persist.\"\n\nStarbucks didn't respond to requests for comment on the study or on baristas' reports of labor practices that are failing to meet the company's stated goals.\n\nIn an internal memo this week published by Time, Starbucks executive Cliff Burrows wrote that the company couldn't validate the survey, but added that \"the findings suggest, contrary to the expectations we have in place, that some partners are receiving their schedules less than one week in advance and that there is a continuing issue with some partners working a close and then an opening shift the following morning.\" He asks store managers \"to go the extra mile to ensure partners have a consistent schedule.\"\n\nStarbucks ends \"race together\" campaign amid public backlash\n\nTo be sure, some Starbucks workers are receiving at least 10 days notice on their work hours, with the survey finding that about one-third receive two weeks notice and another 18 percent get their schedules three weeks in advance. But that leaves almost half of workers who only receive one week's notice, making it more difficult from them manage other obligations, such as school, family commitments or other jobs.\n\nClopens remain a problem, as well. About 60 percent of workers who have to handle a clopen receive seven or fewer hours of rest between a closing and an opening shift, the study found.\n\nThat's prompted one former Starbucks employee to start a petition to end the practice of scheduling clopens. Ciara Moran noted in her petition that she sometimes was only able to get four or five hours of sleep on the days she was scheduled for clopens. She said she quit her job because she doubted whether it was possible to get ahead given the demands on workers.\n\nEven if Starbucks stuck with its policy of providing eight hours between shifts, that's not enough time, especially given that many workers in the service sector have long commutes, the study said.\n\nAnother issue singled out by the report is Starbucks' practices on sick time. Since paid time off is only available to workers with at least a year on the job, about 40 percent of employees in the survey said they had dealt with barriers in taking sick days.\n\nIn a perfect world, Weisse said he'd like to receive his schedule either a month or a", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_title": { - "value": "Kronos Scheduling Algorithm Allegedly Caused Financial Issues for Starbucks Employees", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_description": { - "value": "Kronos’s scheduling algorithm and its use by Starbucks managers allegedly negatively impacted financial and scheduling stability for Starbucks employees, which disadvantaged wage workers.", - "matchLevel": "none", - "matchedWords": [] - } - } - }, - { - "authors": [ - "Marietje Schaake" - ], - "description": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ", - "epoch_date_downloaded": 1661472000, - "epoch_date_modified": 1661731200, - "epoch_date_published": 1661385600, - "epoch_date_submitted": 1661472000, - "image_url": "https://pbs.twimg.com/media/Fa8rYYVWIAAI0xN?format=jpg", - "language": "en", - "report_number": 1967, - "source_domain": "twitter.com", - "submitters": [ - "Cesar Varela" - ], - "title": "Tweet: @MarietjeSchaake", - "name": "Tweet: @MarietjeSchaake", - "url": "https://twitter.com/MarietjeSchaake/status/1562515297688399873", - "tags": [], - "editor_notes": "", - "cloudinary_id": "reports/pbs.twimg.com/media/Fa8rYYVWIAAI0xN?format=jpg", - "text": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague\n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden or the Unabomber, but me… How did that happen? What are Meta’s sources?!\n", - "mongodb_id": "630c8eb443fe03f46cc8bcc4", - "featured": 1, - "is_incident_report": true, - "namespaces": [ - "CSETv1_Annotator-1", - "CSETv1" - ], - "classifications": [ - "CSETv1_Annotator-1:Harm Distribution Basis:none", - "CSETv1_Annotator-1:Sector of Deployment:information and communication", - "CSETv1_Annotator-1:Physical Objects:no", - "CSETv1_Annotator-1:Entertainment Industry:no", - "CSETv1_Annotator-1:Report, Test, or Study of data:no", - "CSETv1_Annotator-1:Deployed:yes", - "CSETv1_Annotator-1:Producer Test in Controlled Conditions:no", - "CSETv1_Annotator-1:Producer Test in Operational Conditions:no", - "CSETv1_Annotator-1:User Test in Controlled Conditions:no", - "CSETv1_Annotator-1:User Test in Operational Conditions:yes", - "CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue", - "CSETv1_Annotator-1:AI System:yes", - "CSETv1_Annotator-1:AI Harm Level:none", - "CSETv1_Annotator-1:Impact on Critical Services:no", - "CSETv1_Annotator-1:Rights Violation:no", - "CSETv1_Annotator-1:Involving Minor:no", - "CSETv1_Annotator-1:Detrimental Content:yes", - "CSETv1_Annotator-1:Protected Characteristic:no", - "CSETv1_Annotator-1:Clear link to Technology:yes", - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:yes", - "CSETv1_Annotator-1:Public Sector Deployment:no", - "CSETv1_Annotator-1:Autonomy Level:Autonomy1", - "CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1_Annotator-1:AI tools and methods:large language models", - "CSETv1_Annotator-1:Special Interest Intangible Harm:yes", - "CSETv1_Annotator-1:Date of Incident Year:2022", - "CSETv1_Annotator-1:Multiple AI Interaction:no", - "CSETv1_Annotator-1:Embedded:no", - "CSETv1_Annotator-1:Location Region:Global", - "CSETv1_Annotator-1:Data Inputs:user prompts", - "CSETv1_Annotator-1:Data Inputs:user queries", - "CSETv1_Annotator-1:Data Inputs:text", - "CSETv1_Annotator-1:AI Task:text generation", - "CSETv1:Harm Distribution Basis:unclear", - "CSETv1:Sector of Deployment:information and communication", - "CSETv1:Physical Objects:no", - "CSETv1:Entertainment Industry:no", - "CSETv1:Report, Test, or Study of data:no", - "CSETv1:Deployed:yes", - "CSETv1:Producer Test in Controlled Conditions:no", - "CSETv1:Producer Test in Operational Conditions:no", - "CSETv1:User Test in Controlled Conditions:no", - "CSETv1:User Test in Operational Conditions:yes", - "CSETv1:Tangible Harm:no tangible harm, near-miss, or issue", - "CSETv1:AI System:yes", - "CSETv1:AI Harm Level:none", - "CSETv1:Impact on Critical Services:no", - "CSETv1:Rights Violation:no", - "CSETv1:Involving Minor:no", - "CSETv1:Detrimental Content:yes", - "CSETv1:Protected Characteristic:no", - "CSETv1:Clear link to Technology:yes", - "CSETv1:Annotator’s AI special interest intangible harm assessment:yes", - "CSETv1:Public Sector Deployment:no", - "CSETv1:Autonomy Level:Autonomy1", - "CSETv1:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1:AI tools and methods:large language models", - "CSETv1:Special Interest Intangible Harm:yes", - "CSETv1:Date of Incident Year:2022", - "CSETv1:Multiple AI Interaction:no", - "CSETv1:Embedded:no", - "CSETv1:Location Region:Global", - "CSETv1:Data Inputs:user prompts", - "CSETv1:Data Inputs:user queries", - "CSETv1:Data Inputs:text", - "CSETv1:AI Task:text generation" - ], - "CSETv1_Annotator-1": { - "Harm Distribution Basis": [ - "none" - ], - "Sector of Deployment": [ - "information and communication" - ], - "Physical Objects": "no", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "no", - "Deployed": "yes", - "Producer Test in Controlled Conditions": "no", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "yes", - "Tangible Harm": "no tangible harm, near-miss, or issue", - "AI System": "yes", - "AI Harm Level": "none", - "Impact on Critical Services": "no", - "Rights Violation": "no", - "Involving Minor": "no", - "Detrimental Content": "yes", - "Protected Characteristic": "no", - "Clear link to Technology": "yes", - "Annotator’s AI special interest intangible harm assessment": "yes", - "Public Sector Deployment": "no", - "Autonomy Level": "Autonomy1", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI tools and methods": [ - "large language models" - ], - "Special Interest Intangible Harm": "yes", - "Date of Incident Year": "2022", - "Multiple AI Interaction": "no", - "Embedded": "no", - "Location Country (two letters)": "", - "Location Region": "Global", - "Infrastructure Sectors": [], - "Operating Conditions": "", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "user prompts", - "user queries", - "text" - ], - "Physical System Type": "", - "AI Task": [ - "text generation" - ] - }, - "CSETv1": { - "Harm Distribution Basis": [ - "unclear" - ], - "Sector of Deployment": [ - "information and communication" - ], - "Physical Objects": "no", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "no", - "Deployed": "yes", - "Producer Test in Controlled Conditions": "no", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "yes", - "Tangible Harm": "no tangible harm, near-miss, or issue", - "AI System": "yes", - "AI Harm Level": "none", - "Impact on Critical Services": "no", - "Rights Violation": "no", - "Involving Minor": "no", - "Detrimental Content": "yes", - "Protected Characteristic": "no", - "Clear link to Technology": "yes", - "Annotator’s AI special interest intangible harm assessment": "yes", - "Public Sector Deployment": "no", - "Autonomy Level": "Autonomy1", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI tools and methods": [ - "large language models" - ], - "Special Interest Intangible Harm": "yes", - "Date of Incident Year": "2022", - "Multiple AI Interaction": "no", - "Embedded": "no", - "Location Country (two letters)": "", - "Location Region": "Global", - "Infrastructure Sectors": [], - "Operating Conditions": "", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "user prompts", - "user queries", - "text" - ], - "Physical System Type": "", - "AI Task": [ - "text generation" - ] - }, - "incident_id": 2, - "incident_date": "2022-08-25", - "epoch_incident_date": 1661385600, - "incident_title": "BlenderBot 3 Cited Dutch Politician as a Terrorist", - "incident_description": "Meta’s conversational AI BlenderBot 3, when prompted “who is a terrorist,“ responded with an incumbent Dutch politician’s name, who was confused about its association.", - "objectID": "1967", - "_snippetResult": { - "description": { - "value": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ", - "matchLevel": "none" - }, - "text": { - "value": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot", - "matchLevel": "none" - } - }, - "_highlightResult": { - "description": { - "value": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ", - "matchLevel": "none", - "matchedWords": [] - }, - "title": { - "value": "Tweet: @MarietjeSchaake", - "matchLevel": "none", - "matchedWords": [] - }, - "text": { - "value": "😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague\n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden or the Unabomber, but me… How did that happen? What are Meta’s sources?!\n", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_title": { - "value": "BlenderBot 3 Cited Dutch Politician as a Terrorist", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_description": { - "value": "Meta’s conversational AI BlenderBot 3, when prompted “who is a terrorist,“ responded with an incumbent Dutch politician’s name, who was confused about its association.", - "matchLevel": "none", - "matchedWords": [] - } - } - }, - { - "authors": [ - "William Douglas Heaven" - ], - "description": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wyn", - "epoch_date_downloaded": 1648598400, - "epoch_date_modified": 1648598400, - "epoch_date_published": 1627603200, - "epoch_date_submitted": 1648598400, - "image_url": "https://wp.technologyreview.com/wp-content/uploads/2021/07/AP_20178810286930.jpg?resize=1200,600", - "language": "en", - "report_number": 1551, - "source_domain": "technologyreview.com", - "submitters": [ - "Anonymous" - ], - "title": "Hundreds of AI tools have been built to catch covid. None of them helped.", - "name": "Hundreds of AI tools have been built to catch covid. None of them helped.", - "url": "https://www.technologyreview.com/2021/07/30/1030329/machine-learning-ai-failed-covid-hospital-diagnosis-pandemic/", - "tags": [], - "editor_notes": "", - "cloudinary_id": "reports/wp.technologyreview.com/wp-content/uploads/2021/07/AP_20178810286930.jpg?resize=1200,600", - "text": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wynants, an epidemiologist at Maastricht University in the Netherlands, who studies predictive tools.\n\nBut there was data coming out of China, which had a four-month head start in the race to beat the pandemic. If machine-learning algorithms could be trained on that data to help doctors understand what they were seeing and make decisions, it just might save lives. “I thought, ‘If there’s any time that AI could prove its usefulness, it’s now,’” says Wynants. “I had my hopes up.”\n\nIt never happened—but not for lack of effort. Research teams around the world stepped up to help. The AI community, in particular, rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines—in theory.\n\nIn the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful.\n\nThat’s the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK’s national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid.\n\nNot fit for clinical use\n\nThis echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Wynants is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing.\n\n“It’s shocking,” says Wynants. “I went into it with some worries, but this exceeded my fears.”\n\nWynants’s study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use.\n\n“This pandemic was a big test for AI and medicine,” says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. “It would have gone a long way to getting the public on our side,” he says. “But I don’t think we passed that test.”\n\nBoth teams found that researchers repeated the same basic errors in the way they trained or tested their tools. Incorrect assumptions about the data often meant that the trained models did not work as claimed.\n\nWynants and Driggs still believe AI has the potential to help. But they are concerned that it could be harmful if built in the wrong way because they could miss diagnoses or underestimate risk for vulnerable patients. “There is a lot of hype about machine-learning models and what they can do today,” says Driggs.\n\nUnrealistic expectations encourage the use of these tools before they are ready. Wynants and Driggs both say that a few of the algorithms they looked at have already been used in hospitals, and some are being marketed by private developers. “I fear that they may have harmed patients,” says Wynants.\n\nSo what went wrong? And how do we bridge that gap? If there’s an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. “The pandemic has put problems in the spotlight that we’ve been dragging along for some time,” says Wynants.\n\nWhat went wrong\n\nMany of the problems that were uncovered are linked to the poor quality of the data that researchers used to develop their tools. Information about covid patients, including medical scans, was collected and shared in the middle of a global pandemic, often by the doctors struggling to treat those patients. Researchers wanted to help quickly, and these were the only public data sets available. But this meant that many tools were built using mislabeled data or data from unknown sources.\n\nDriggs highlights the problem of what he calls Frankenstein data sets, which are spliced together from multiple sources and can contain duplicates. This means that some tools end up being tested on the same data they were trained on, making them appear more accurate than they are.\n\nIt also muddies the origin of certain data sets. This can mean that researchers miss important features that skew the training of their models. Many unwittingly used a data set that contained chest scans of children who did not have covid as their examples of what non-covid cases looked like. But as a result, the AIs learned to identify kids, not covid.\n\nDriggs’s group trained its own model using a data set that contained a mix of scans taken when patients were lying down and standing up. Because patients scanned while lying down were more likely to be seriously ill, the AI learned wrongly to predict serious covid risk from a person’s position.\n\nIn yet other cases, some AIs were found to be picking up on the text font that certain hospitals used to label the scans. As a result, fonts from hospitals with more serious caseloads became predictors of covid risk.\n\nErrors like these seem obvious in hindsight. They can also be fixed by adjusting the models, if researchers are aware of them. It is possible to acknowledge the shortcomings and release a less accurate, but less misleading model. But many tools were developed either by AI researchers who lacked the medical expertise to spot flaws in the data or by medical researchers who lacked the mathematical skills to compensate for those flaws.\n\nA more subtle problem Driggs highlights is incorporation bias, or bias introduced at the point a data set is labeled. For example, many medical scans were labeled according to whether the radiologists who created them said they showed covid. But that embeds, or incorporates, any biases of that particular doctor into the ground truth of a data set. It would be much better to label a medical scan with the result of a PCR test rather than one doctor’s opinion, says Driggs. But there isn’t always time for statistical niceties in busy hospitals.\n\nThat hasn’t stopped some of these tools from being rushed into clinical practice. Wynants says it isn’t clear which ones are being used or how. Hospitals will sometimes say that they are using a tool only for research purposes, which makes it hard to assess how much doctors are relying on them. “There’s a lot of secrecy,” she says.\n\nWynants asked one company that was marketing deep-learning algorithms to share information about its approach but did not hear back. She later found several published models from researchers tied to this company, all of them with a high risk of bias. “We don’t actually know what the company implemented,” she says.\n\nAccording to Wynants, some hospitals are even signing nondisclosure agreements with medical AI vendors. When she asked doctors what algorithms or software they were using, they sometimes told her they weren’t allowed to say.\n\nHow to fix it\n\nWhat’s the fix? Better data would help, but in times of crisis that’s a big ask. It’s more important to make the most of the data sets we have. The simplest move would be for AI teams to collaborate more with clinicians, says Driggs. Researchers also need to share their models and disclose how they were trained so that others can test them and build on them. “Those are two things we could do today,” he says. “And they would solve mayb", - "mongodb_id": "6243a9eedf8b4b62d982817e", - "featured": 1, - "is_incident_report": true, - "namespaces": [ - "CSETv1_Annotator-3", - "CSETv1_Annotator-1" - ], - "classifications": [ - "CSETv1_Annotator-3:Harm Distribution Basis:unclear", - "CSETv1_Annotator-3:Sector of Deployment:human health and social work activities", - "CSETv1_Annotator-3:Physical Objects:yes", - "CSETv1_Annotator-3:Entertainment Industry:no", - "CSETv1_Annotator-3:Report, Test, or Study of data:no", - "CSETv1_Annotator-3:Deployed:yes", - "CSETv1_Annotator-3:Producer Test in Controlled Conditions:no", - "CSETv1_Annotator-3:Producer Test in Operational Conditions:no", - "CSETv1_Annotator-3:User Test in Controlled Conditions:no", - "CSETv1_Annotator-3:User Test in Operational Conditions:no", - "CSETv1_Annotator-3:Tangible Harm:unclear", - "CSETv1_Annotator-3:AI System:yes", - "CSETv1_Annotator-3:AI Harm Level:none", - "CSETv1_Annotator-3:Impact on Critical Services:maybe", - "CSETv1_Annotator-3:Rights Violation:maybe", - "CSETv1_Annotator-3:Involving Minor:no", - "CSETv1_Annotator-3:Detrimental Content:no", - "CSETv1_Annotator-3:Protected Characteristic:maybe", - "CSETv1_Annotator-3:Clear link to Technology:yes", - "CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:maybe", - "CSETv1_Annotator-3:Public Sector Deployment:no", - "CSETv1_Annotator-3:Autonomy Level:unclear", - "CSETv1_Annotator-3:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1_Annotator-3:Special Interest Intangible Harm:maybe", - "CSETv1_Annotator-3:Date of Incident Year:2020", - "CSETv1_Annotator-3:Multiple AI Interaction:no", - "CSETv1_Annotator-3:Embedded:maybe", - "CSETv1_Annotator-3:Location Region:Global", - "CSETv1_Annotator-3:Infrastructure Sectors:healthcare and public health", - "CSETv1_Annotator-3:Data Inputs:Unclear", - "CSETv1_Annotator-1:Physical Objects:no", - "CSETv1_Annotator-1:Entertainment Industry:no", - "CSETv1_Annotator-1:Report, Test, or Study of data:maybe", - "CSETv1_Annotator-1:Deployed:no", - "CSETv1_Annotator-1:Producer Test in Controlled Conditions:yes", - "CSETv1_Annotator-1:Producer Test in Operational Conditions:no", - "CSETv1_Annotator-1:User Test in Controlled Conditions:no", - "CSETv1_Annotator-1:User Test in Operational Conditions:no", - "CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue", - "CSETv1_Annotator-1:AI System:yes", - "CSETv1_Annotator-1:AI Harm Level:none", - "CSETv1_Annotator-1:Impact on Critical Services:no", - "CSETv1_Annotator-1:Rights Violation:no", - "CSETv1_Annotator-1:Involving Minor:no", - "CSETv1_Annotator-1:Detrimental Content:no", - "CSETv1_Annotator-1:Protected Characteristic:no", - "CSETv1_Annotator-1:Harm Distribution Basis:none", - "CSETv1_Annotator-1:Special Interest Intangible Harm:no", - "CSETv1_Annotator-1:Clear link to Technology:maybe", - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no", - "CSETv1_Annotator-1:Date of Incident Year:2020", - "CSETv1_Annotator-1:Multiple AI Interaction:no", - "CSETv1_Annotator-1:Embedded:no", - "CSETv1_Annotator-1:Location Region:Global", - "CSETv1_Annotator-1:Data Inputs:patient information", - "CSETv1_Annotator-1:Data Inputs:medical scans", - "CSETv1_Annotator-1:Data Inputs:chest scans", - "CSETv1_Annotator-1:Data Inputs:electronic health records", - "CSETv1_Annotator-1:Sector of Deployment:human health and social work activities", - "CSETv1_Annotator-1:Public Sector Deployment:no", - "CSETv1_Annotator-1:Autonomy Level:Autonomy3", - "CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm", - "CSETv1_Annotator-1:AI Task:diagnose patients" - ], - "CSETv1_Annotator-3": { - "Harm Distribution Basis": [ - "unclear" - ], - "Sector of Deployment": [ - "human health and social work activities" - ], - "Physical Objects": "yes", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "no", - "Deployed": "yes", - "Producer Test in Controlled Conditions": "no", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "no", - "Tangible Harm": "unclear", - "AI System": "yes", - "AI Harm Level": "none", - "Impact on Critical Services": "maybe", - "Rights Violation": "maybe", - "Involving Minor": "no", - "Detrimental Content": "no", - "Protected Characteristic": "maybe", - "Clear link to Technology": "yes", - "Annotator’s AI special interest intangible harm assessment": "maybe", - "Public Sector Deployment": "no", - "Autonomy Level": "unclear", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI tools and methods": "", - "Special Interest Intangible Harm": "maybe", - "Date of Incident Year": "2020", - "Multiple AI Interaction": "no", - "Embedded": "maybe", - "Location Country (two letters)": "", - "Location Region": "Global", - "Infrastructure Sectors": [ - "healthcare and public health" - ], - "Operating Conditions": "", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "Unclear" - ], - "Physical System Type": "", - "AI Task": "" - }, - "CSETv1_Annotator-1": { - "Physical Objects": "no", - "Entertainment Industry": "no", - "Report, Test, or Study of data": "maybe", - "Deployed": "no", - "Producer Test in Controlled Conditions": "yes", - "Producer Test in Operational Conditions": "no", - "User Test in Controlled Conditions": "no", - "User Test in Operational Conditions": "no", - "Tangible Harm": "no tangible harm, near-miss, or issue", - "AI System": "yes", - "AI Harm Level": "none", - "Impact on Critical Services": "no", - "Rights Violation": "no", - "Involving Minor": "no", - "Detrimental Content": "no", - "Protected Characteristic": "no", - "Harm Distribution Basis": [ - "none" - ], - "Special Interest Intangible Harm": "no", - "Clear link to Technology": "maybe", - "Annotator’s AI special interest intangible harm assessment": "no", - "Date of Incident Year": "2020", - "Multiple AI Interaction": "no", - "Embedded": "no", - "Location Region": "Global", - "Lives Lost": 0, - "Injuries": 0, - "Data Inputs": [ - "patient information", - "medical scans", - "chest scans", - "electronic health records" - ], - "Sector of Deployment": [ - "human health and social work activities" - ], - "Public Sector Deployment": "no", - "Autonomy Level": "Autonomy3", - "Intentional Harm": "No. Not intentionally designed to perform harm", - "AI Task": [ - "diagnose patients" - ] - }, - "incident_id": 3, - "incident_date": "2021-07-30", - "epoch_incident_date": 1627603200, - "incident_title": "AI Tools Failed to Sufficiently Predict COVID Patients, Some Potentially Harmful", - "incident_description": "AI tools failed to sufficiently predict COVID patients, some potentially harmful.", - "objectID": "1551", - "_snippetResult": { - "description": { - "value": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wyn", - "matchLevel": "none" - }, - "text": { - "value": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis", - "matchLevel": "none" - } - }, - "_highlightResult": { - "description": { - "value": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wyn", - "matchLevel": "none", - "matchedWords": [] - }, - "title": { - "value": "Hundreds of AI tools have been built to catch covid. None of them helped.", - "matchLevel": "none", - "matchedWords": [] - }, - "text": { - "value": "When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wynants, an epidemiologist at Maastricht University in the Netherlands, who studies predictive tools.\n\nBut there was data coming out of China, which had a four-month head start in the race to beat the pandemic. If machine-learning algorithms could be trained on that data to help doctors understand what they were seeing and make decisions, it just might save lives. “I thought, ‘If there’s any time that AI could prove its usefulness, it’s now,’” says Wynants. “I had my hopes up.”\n\nIt never happened—but not for lack of effort. Research teams around the world stepped up to help. The AI community, in particular, rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines—in theory.\n\nIn the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful.\n\nThat’s the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK’s national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid.\n\nNot fit for clinical use\n\nThis echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Wynants is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing.\n\n“It’s shocking,” says Wynants. “I went into it with some worries, but this exceeded my fears.”\n\nWynants’s study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use.\n\n“This pandemic was a big test for AI and medicine,” says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. “It would have gone a long way to getting the public on our side,” he says. “But I don’t think we passed that test.”\n\nBoth teams found that researchers repeated the same basic errors in the way they trained or tested their tools. Incorrect assumptions about the data often meant that the trained models did not work as claimed.\n\nWynants and Driggs still believe AI has the potential to help. But they are concerned that it could be harmful if built in the wrong way because they could miss diagnoses or underestimate risk for vulnerable patients. “There is a lot of hype about machine-learning models and what they can do today,” says Driggs.\n\nUnrealistic expectations encourage the use of these tools before they are ready. Wynants and Driggs both say that a few of the algorithms they looked at have already been used in hospitals, and some are being marketed by private developers. “I fear that they may have harmed patients,” says Wynants.\n\nSo what went wrong? And how do we bridge that gap? If there’s an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. “The pandemic has put problems in the spotlight that we’ve been dragging along for some time,” says Wynants.\n\nWhat went wrong\n\nMany of the problems that were uncovered are linked to the poor quality of the data that researchers used to develop their tools. Information about covid patients, including medical scans, was collected and shared in the middle of a global pandemic, often by the doctors struggling to treat those patients. Researchers wanted to help quickly, and these were the only public data sets available. But this meant that many tools were built using mislabeled data or data from unknown sources.\n\nDriggs highlights the problem of what he calls Frankenstein data sets, which are spliced together from multiple sources and can contain duplicates. This means that some tools end up being tested on the same data they were trained on, making them appear more accurate than they are.\n\nIt also muddies the origin of certain data sets. This can mean that researchers miss important features that skew the training of their models. Many unwittingly used a data set that contained chest scans of children who did not have covid as their examples of what non-covid cases looked like. But as a result, the AIs learned to identify kids, not covid.\n\nDriggs’s group trained its own model using a data set that contained a mix of scans taken when patients were lying down and standing up. Because patients scanned while lying down were more likely to be seriously ill, the AI learned wrongly to predict serious covid risk from a person’s position.\n\nIn yet other cases, some AIs were found to be picking up on the text font that certain hospitals used to label the scans. As a result, fonts from hospitals with more serious caseloads became predictors of covid risk.\n\nErrors like these seem obvious in hindsight. They can also be fixed by adjusting the models, if researchers are aware of them. It is possible to acknowledge the shortcomings and release a less accurate, but less misleading model. But many tools were developed either by AI researchers who lacked the medical expertise to spot flaws in the data or by medical researchers who lacked the mathematical skills to compensate for those flaws.\n\nA more subtle problem Driggs highlights is incorporation bias, or bias introduced at the point a data set is labeled. For example, many medical scans were labeled according to whether the radiologists who created them said they showed covid. But that embeds, or incorporates, any biases of that particular doctor into the ground truth of a data set. It would be much better to label a medical scan with the result of a PCR test rather than one doctor’s opinion, says Driggs. But there isn’t always time for statistical niceties in busy hospitals.\n\nThat hasn’t stopped some of these tools from being rushed into clinical practice. Wynants says it isn’t clear which ones are being used or how. Hospitals will sometimes say that they are using a tool only for research purposes, which makes it hard to assess how much doctors are relying on them. “There’s a lot of secrecy,” she says.\n\nWynants asked one company that was marketing deep-learning algorithms to share information about its approach but did not hear back. She later found several published models from researchers tied to this company, all of them with a high risk of bias. “We don’t actually know what the company implemented,” she says.\n\nAccording to Wynants, some hospitals are even signing nondisclosure agreements with medical AI vendors. When she asked doctors what algorithms or software they were using, they sometimes told her they weren’t allowed to say.\n\nHow to fix it\n\nWhat’s the fix? Better data would help, but in times of crisis that’s a big ask. It’s more important to make the most of the data sets we have. The simplest move would be for AI teams to collaborate more with clinicians, says Driggs. Researchers also need to share their models and disclose how they were trained so that others can test them and build on them. “Those are two things we could do today,” he says. “And they would solve mayb", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_title": { - "value": "AI Tools Failed to Sufficiently Predict COVID Patients, Some Potentially Harmful", - "matchLevel": "none", - "matchedWords": [] - }, - "incident_description": { - "value": "AI tools failed to sufficiently predict COVID patients, some potentially harmful.", - "matchLevel": "none", - "matchedWords": [] - } - } - }, + authors: ['Aimee Picchi'], + description: + "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", + epoch_date_downloaded: 1555113600, + epoch_date_modified: 1725431608, + epoch_date_published: 1443052800, + epoch_date_submitted: 1559347200, + image_url: + 'https://cbsnews1.cbsistatic.com/hub/i/r/2015/03/17/01a38576-5108-40f7-8df8-5416164ed878/thumbnail/1200x630/ca8d35fe6bc065b5c9a747d92bc6d94c/154211248.jpg', + language: 'en', + report_number: 23, + source_domain: 'cbsnews.com', + submitters: ['Catherine Olsson'], + title: '​Is Starbucks shortchanging its baristas?', + name: '​Is Starbucks shortchanging its baristas?', + url: 'https://www.cbsnews.com/news/is-starbucks-shortchanging-its-baristas/', + tags: [], + editor_notes: '', + cloudinary_id: + 'reports/cbsnews1.cbsistatic.com/hub/i/r/2015/03/17/01a38576-5108-40f7-8df8-5416164ed878/thumbnail/1200x630/ca8d35fe6bc065b5c9a747d92bc6d94c/154211248.jpg', + text: "For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health insurance and some extra money while he studies at Georgia Perimeter College. What it doesn't provide is the kind of stable schedule that the company promised its workers last year.\n\n\"It's the wild inconsistency\" of the hours that's a problem, Weisse, 32, said. \"We're supposed to get them 10 days in advance, which often happens, but there's no guarantee. If our manager doesn't get it to us on time, we just have to deal with it.\"\n\nThat became a problem recently when Weisse's manager gave him only a few days notice on his work hours, which ended up conflicting with an anatomy and physiology exam at his college. Weisse ended up paying another worker $20 to take his shift so he could take the exam.\n\nThe short notice is especially frustrating because of Starbucks' vow last year to post employees' schedules at least 10 days in advance, as well as the company's insistence that workers provide at least one-month notice when they need to take a day off.\n\nWhat's behind Starbucks price increases?\n\nWeisse isn't alone in complaining that Starbucks isn't living up to its promises to overhaul its labor practices for its roughly 130,000 baristas. That vow followed an article last year by The New York Times that detailed how workers were struggling to manage childcare and other obligations when the company provided only a few days notice about their schedules.\n\nAbout half of roughly 200 Starbucks baristas in a recent survey said they are still receiving their schedule with less than one week's notice. Others also reported being asked to handle \"clopens,\" split shifts in which employees work a closing shift late into the evening and then an early opening shift the following morning. The company last year promised to end the practice.\n\nOf course, Starbucks isn't alone in using \"just-in-time\" scheduling, with the retail and restaurant industry increasingly turning to software that allows them to change work schedules at the last minute, depending on whether business picks up or slows down. But it is Starbucks that has become a lightning rod on the issue given its vows to improve how it treats employees and its own emphatic claims to valuing workers, whom it labels \"partners.\"\n\n\"Starbucks has the values and wants to do right by their employees,\" said Carrie Gleason, director of the Fair Workweek Initiative at the Center for Popular Democracy, an advocacy group focused on workers'rights, and a co-author of the group's new report on the company's labor practices. \"However, since last year when the company recognized there was a serious problem with the way it scheduled workers and pledged to reform, still so many of the same issues persist.\"\n\nStarbucks didn't respond to requests for comment on the study or on baristas' reports of labor practices that are failing to meet the company's stated goals.\n\nIn an internal memo this week published by Time, Starbucks executive Cliff Burrows wrote that the company couldn't validate the survey, but added that \"the findings suggest, contrary to the expectations we have in place, that some partners are receiving their schedules less than one week in advance and that there is a continuing issue with some partners working a close and then an opening shift the following morning.\" He asks store managers \"to go the extra mile to ensure partners have a consistent schedule.\"\n\nStarbucks ends \"race together\" campaign amid public backlash\n\nTo be sure, some Starbucks workers are receiving at least 10 days notice on their work hours, with the survey finding that about one-third receive two weeks notice and another 18 percent get their schedules three weeks in advance. But that leaves almost half of workers who only receive one week's notice, making it more difficult from them manage other obligations, such as school, family commitments or other jobs.\n\nClopens remain a problem, as well. About 60 percent of workers who have to handle a clopen receive seven or fewer hours of rest between a closing and an opening shift, the study found.\n\nThat's prompted one former Starbucks employee to start a petition to end the practice of scheduling clopens. Ciara Moran noted in her petition that she sometimes was only able to get four or five hours of sleep on the days she was scheduled for clopens. She said she quit her job because she doubted whether it was possible to get ahead given the demands on workers.\n\nEven if Starbucks stuck with its policy of providing eight hours between shifts, that's not enough time, especially given that many workers in the service sector have long commutes, the study said.\n\nAnother issue singled out by the report is Starbucks' practices on sick time. Since paid time off is only available to workers with at least a year on the job, about 40 percent of employees in the survey said they had dealt with barriers in taking sick days.\n\nIn a perfect world, Weisse said he'd like to receive his schedule either a month or a", + mongodb_id: '5d34b8c29ced494f010ed470', + featured: 2, + flag: true, + is_incident_report: true, + is_translated: false, + namespaces: ['CSETv0', 'GMF', 'CSETv1_Annotator-1', 'CSETv1'], + classifications: [ + 'CSETv0:Location:Global', + 'CSETv0:Near Miss:Unclear/unknown', + 'CSETv0:Named Entities:Starbucks', + 'CSETv0:Named Entities:Kronos', + 'CSETv0:Technology Purveyor:Starbucks', + 'CSETv0:Intent:Unclear', + 'CSETv0:Severity:Negligible', + 'CSETv0:Harm Type:Psychological harm', + 'CSETv0:Infrastructure Sectors:Food and agriculture', + 'GMF:Known AI Technical Failure:Underspecification', + 'GMF:Potential AI Technical Failure:Tuning Issues', + 'GMF:Potential AI Technical Failure:Misconfigured Aggregation', + 'GMF:Potential AI Technology:Regression', + 'GMF:Potential AI Technology:Diverse Data', + 'GMF:Known AI Goal:Market Forecasting', + 'GMF:Known AI Goal:Scheduling', + 'CSETv1_Annotator-1:Physical Objects:no', + 'CSETv1_Annotator-1:Entertainment Industry:no', + 'CSETv1_Annotator-1:Report, Test, or Study of data:no', + 'CSETv1_Annotator-1:Deployed:yes', + 'CSETv1_Annotator-1:Producer Test in Controlled Conditions:no', + 'CSETv1_Annotator-1:Producer Test in Operational Conditions:no', + 'CSETv1_Annotator-1:User Test in Controlled Conditions:no', + 'CSETv1_Annotator-1:User Test in Operational Conditions:no', + 'CSETv1_Annotator-1:Tangible Harm:tangible harm definitively occurred', + 'CSETv1_Annotator-1:AI System:yes', + 'CSETv1_Annotator-1:AI Harm Level:AI tangible harm event', + 'CSETv1_Annotator-1:Impact on Critical Services:no', + 'CSETv1_Annotator-1:Rights Violation:no', + 'CSETv1_Annotator-1:Involving Minor:no', + 'CSETv1_Annotator-1:Detrimental Content:no', + 'CSETv1_Annotator-1:Protected Characteristic:no', + 'CSETv1_Annotator-1:Harm Distribution Basis:none', + 'CSETv1_Annotator-1:Clear link to Technology:yes', + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no', + 'CSETv1_Annotator-1:Sector of Deployment:wholesale and retail trade', + 'CSETv1_Annotator-1:Sector of Deployment:accommodation and food service activities', + 'CSETv1_Annotator-1:Public Sector Deployment:no', + 'CSETv1_Annotator-1:Autonomy Level:Autonomy2', + 'CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1_Annotator-1:Special Interest Intangible Harm:no', + 'CSETv1_Annotator-1:Date of Incident Year:2014', + 'CSETv1_Annotator-1:Multiple AI Interaction:no', + 'CSETv1_Annotator-1:Embedded:no', + 'CSETv1_Annotator-1:Location Country (two letters):US', + 'CSETv1_Annotator-1:Location Region:North America', + 'CSETv1_Annotator-1:Data Inputs:schedules', + 'CSETv1_Annotator-1:Data Inputs:worker profiles', + 'CSETv1_Annotator-1:AI Task:scheduling', + 'CSETv1_Annotator-1:AI Task:productivity optimization', + 'CSETv1:Harm Distribution Basis:none', + 'CSETv1:Sector of Deployment:accommodation and food service activities', + 'CSETv1:Physical Objects:no', + 'CSETv1:Entertainment Industry:no', + 'CSETv1:Report, Test, or Study of data:no', + 'CSETv1:Deployed:yes', + 'CSETv1:Producer Test in Controlled Conditions:no', + 'CSETv1:Producer Test in Operational Conditions:no', + 'CSETv1:User Test in Controlled Conditions:no', + 'CSETv1:User Test in Operational Conditions:no', + 'CSETv1:Tangible Harm:tangible harm definitively occurred', + 'CSETv1:AI System:yes', + 'CSETv1:AI Harm Level:AI tangible harm event', + 'CSETv1:Impact on Critical Services:no', + 'CSETv1:Rights Violation:no', + 'CSETv1:Involving Minor:no', + 'CSETv1:Detrimental Content:no', + 'CSETv1:Protected Characteristic:no', + 'CSETv1:Clear link to Technology:yes', + 'CSETv1:Annotator’s AI special interest intangible harm assessment:no', + 'CSETv1:Public Sector Deployment:no', + 'CSETv1:Autonomy Level:Autonomy2', + 'CSETv1:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1:Special Interest Intangible Harm:no', + 'CSETv1:Date of Incident Year:2014', + 'CSETv1:Multiple AI Interaction:no', + 'CSETv1:Embedded:no', + 'CSETv1:Location Country (two letters):US', + 'CSETv1:Location Region:North America', + 'CSETv1:Data Inputs:schedules', + 'CSETv1:Data Inputs:worker profiles', + 'CSETv1:Data Inputs:store traffic', + 'CSETv1:AI Task:scheduling', + 'CSETv1:AI Task:productivity optimization', + 'CSETv1:AI Task:predict store traffic', + ], + CSETv0: { + Location: 'Global', + 'Near Miss': 'Unclear/unknown', + 'Named Entities': ['Starbucks', 'Kronos'], + 'Technology Purveyor': ['Starbucks'], + Intent: 'Unclear', + Severity: 'Negligible', + 'Harm Type': ['Psychological harm'], + 'Lives Lost': false, + 'Harm Distribution Basis': [], + 'Infrastructure Sectors': ['Food and agriculture'], + 'Financial Cost': '', + 'System Developer': [], + 'Sector of Deployment': [''], + 'Public Sector Deployment': false, + 'Nature of End User': '', + 'Level of Autonomy': '', + 'Relevant AI functions': [], + 'AI Techniques': [], + 'AI Applications': [], + 'Physical System': [], + 'Problem Nature': [], + }, + GMF: { + 'Known AI Technical Failure': ['Underspecification'], + 'Known AI Technical Failure Snippets': [ + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"“You’re waiting on your job to control your life,” she said, with the scheduling software used by her employer dictating everything from “how much sleep Gavin will get to what groceries I’ll be able to buy this month.”"', + }, + { + short_name: 'Related Classifications', + value_json: '["Underspecification"]', + }, + ], + }, + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"Along with virtually every major retail and restaurant chain, Starbucks relies on software that choreographs workers in precise, intricate ballets, using sales patterns and other data to determine which of its 130,000 baristas are needed in its thousands of locations and exactly when."', + }, + { + short_name: 'Related Classifications', + value_json: '["Underspecification"]', + }, + ], + }, + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"Among other changes, the company said it would end the practice of \\"clopening,\\" when an employee responsible for closing a store late at night is also assigned to open it early in the morning."', + }, + { + short_name: 'Related Classifications', + value_json: '["Underspecification"]', + }, + ], + }, + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"In a follow-up piece, the author, Jodi Kantor, points directly to Kronos\' scheduling software as the root of the problem."', + }, + { + short_name: 'Related Classifications', + value_json: '["Underspecification"]', + }, + ], + }, ], - "nbHits": 3, - "page": 0, - "nbPages": 1, - "hitsPerPage": 28, - "facets": { - "flag": { - "true": 36 - }, - "tags": { - "response": 163, - "deepfake": 23, - "generative AI": 12, - "CSAM": 5, - "journalism": 5, - "Amazon": 4, - "minors": 4, - "racial bias": 4, - "Audio": 3, - "Deepfakes": 3, - "Grok": 3, - "Scam": 3, - "TikTok": 3, - "deepfake porn": 3, - "fraud": 3, - "social media": 3, - "Alexa": 2, - "Celebrities": 2, - "Children": 2, - "Data privacy": 2, - "Education": 2, - "Extortion": 2, - "Facial recognition": 2, - "Kamala Harris": 2, - "Misinformation": 2, - "Nazi": 2, - "Nudify": 2, - "Privacy": 2, - "algorithmic bias": 2, - "chatbot": 2, - "child abuse": 2, - "child sex abuse material": 2, - "content moderation": 2, - "data": 2, - "driverless cars": 2, - "elections": 2, - "facial recognition technology": 2, - "health insurance": 2, - "lack of consent": 2, - "products": 2, - "#response": 1, - "2FA": 1, - "AI Music": 1, - "AI bias": 1, - "AI content farms": 1, - "AI facial feature modification": 1, - "AI image alteration": 1, - "AI meal planner": 1, - "AI misinterpretation of professionalism": 1, - "AI racial representation": 1, - "AI research": 1, - "AI-authored books": 1, - "AI-generated news sites": 1, - "Advertisement": 1, - "Alaska": 1, - "Amazon Photos": 1, - "Apple Photos": 1, - "Army": 1, - "Barack Obama": 1, - "British politics": 1, - "California law": 1, - "Central America": 1, - "ChatGPT": 1, - "Chatbots": 1, - "Cheating": 1, - "Child sexual abuse material": 1, - "Clearview AI": 1, - "Criminal activity": 1, - "Crypto": 1, - "DPA": 1, - "Dark Web": 1, - "Data exposure": 1, - "Data retention": 1, - "Data set": 1, - "Donald Trump": 1, - "Ed Tech": 1, - "Educational AI": 1, - "Election interference": 1, - "Elon Musk": 1, - "Encouragement of violence": 1, - "Exam": 1, - "FTC": 1, - "Fake news": 1, - "False positive": 1, - "Figma": 1, - "GitHub": 1, - "Google Photos": 1, - "IWF": 1, - "Image recognition": 1, - "India": 1, - "Internet Watch Foundation": 1, - "LAION-5B": 1, - "LGBTQ+": 1, - "Latin America": 1, - "Lawsuit": 1, - "Mental illness": 1, - "Microsoft Azure": 1, - "Microsoft OneDrive": 1, - "Music": 1, - "Nudification": 1, - "OSHA": 1, - "Photoshop": 1, - "Plagiarism detector": 1, - "Political manipulation": 1, - "Porn": 1, - "Queer representation": 1, - "Robots": 1, - "Russia": 1, - "SAS tokens": 1, - "Security vulnerability": 1, - "Slovakia": 1, - "South America": 1, - "South Korea": 1, - "Spain": 1, - "Spam": 1, - "Stable Diffusion": 1, - "Tesla": 1, - "Training data": 1, - "Turkish": 1, - "Twitter": 1, - "Undress": 1, - "Video": 1, - "Workplace accident": 1, - "X": 1, - "accountability": 1, - "art": 1, - "automated decision-making": 1, - "autopilot": 1, - "biometric data": 1, - "celebrity": 1, - "child exploitation": 1, - "child pornography": 1, - "comedy": 1, - "commercials": 1, - "conference call": 1, - "copyright infringement": 1, - "court": 1, - "coverage denial": 1, - "customer service": 1, - "data accuracy": 1, - "data integrity": 1, - "deepfake audio": 1, - "deepfaked audio": 1, - "digital crime": 1, - "disinformation": 1, - "editing error": 1, - "false arrest": 1, - "false citations": 1, - "gender bias": 1, - "generated books": 1, - "geopolitics": 1, - "google ai search": 1, - "hallucination": 1, - "hazardous advice": 1, - "healthcare algorithms": 1, - "healthcare modeling": 1, - "healthcare technology": 1, - "inappropriate content": 1, - "influence campaigns": 1, - "journalism ethics": 1, - "labor relations": 1, - "language model malfunction": 1, - "law": 1, - "legal": 1, - "legal dispute": 1, - "loans": 1, - "medical claims": 1, - "medical diagnosis": 1, - "misidentification": 1, - "news aggregation": 1, - "news network": 1, - "offensive content": 1, - "official report using false generated AI": 1, - "pedestrians": 1, - "photo editing": 1, - "political disinformation": 1, - "politics": 1, - "posthumous use": 1, - "retail": 1, - "robotics": 1, - "self-driving": 1, - "synthetic imagery": 1, - "targeted advertising": 1, - "text-to-image generation": 1, - "underwriting": 1, - "website poll": 1, - "wifi connectivity": 1, - "workplace death": 1, - "wrongful arrest": 1 - }, - "authors": { - "Associated Press": 25, - "Reuters": 23, - "BBC News": 22, - "James Vincent": 22, - "Kashmir Hill": 18, - "Christopher Knaus": 13, - "Katyanna Quach": 13, - "Benj Edwards": 12, - "Noor Al-Sibai": 12, - "Alex Hern": 11, - "Andrew J. Hawkins": 11, - "Fred Lambert": 11, - "Tom Simonite": 11, - "Emma Roth": 10, - "Maggie Harrison": 10, - "Will Oremus": 10, - "Chloe Xiang": 9, - "Pranshu Verma": 9, - "Sam Levin": 9, - "Surya Mattu": 9, - "Bloomberg": 8, - "Dan Milmo": 8, - "David Shepardson": 8, - "Drew Harwell": 8, - "James Whitbrook": 8, - "Julia Angwin": 8, - "Karen Hao": 8, - "The Guardian": 8, - "Tom Krisher": 8, - "Aarian Marshall": 7, - "Ashley Belanger": 7, - "Jon Fingas": 7, - "Kat Tenbarge": 7, - "Mack DeGeurin": 7, - "Matthew Gault": 7, - "Pesala Bandara": 7, - "Samantha Cole": 7, - "Steven Lee Myers": 7, - "Agence France-Presse": 6, - "Gerrit De Vynck": 6, - "Jon Christian": 6, - "Kathleen McGrory": 6, - "Matt McFarland": 6, - "Neal E. Boudette": 6, - "Olivia Solon": 6, - "Ryan Mac": 6, - "Samuel Gibbs": 6, - "Sky News": 6, - "Tim Cushing": 6, - "Todd Feathers": 6, - "Will Knight": 6, - "Brandon Vigliarolo": 5, - "Britney Nguyen": 5, - "Cade Metz": 5, - "Dan Goodin": 5, - "Faiz Siddiqui": 5, - "Garance Burke": 5, - "Human Rights Watch": 5, - "James Titcomb": 5, - "Jason Koebler": 5, - "Jay Peters": 5, - "John Goreham": 5, - "Josh Taylor": 5, - "Julie Jargon": 5, - "Lora Kolodny": 5, - "Melissa Heikkilä": 5, - "Natasha Lomas": 5, - "Natasha Singer": 5, - "Nico Grant": 5, - "Nitasha Tiku": 5, - "Ryan Felton": 5, - "Shannon Bond": 5, - "Stuart A. Thompson": 5, - "The Economist": 5, - "Thomas Claburn": 5, - "Tiffany Hsu": 5, - "Timothy B. Lee": 5, - "Trevor Mogg": 5, - "Victor Tangermann": 5, - "Wikipedia Editors": 5, - "Afp": 4, - "Arden Dier": 4, - "BBC": 4, - "Benjamin Weiser": 4, - "Brad Templeton": 4, - "CBS News": 4, - "Casey Ross": 4, - "Cyrus Farivar": 4, - "Dave Lee": 4, - "David Meyer": 4, - "Derek B. Johnson": 4, - "Elizabeth Stanton": 4, - "Federal Trade Commission": 4, - "Gabriel Geiger": 4, - "Glenn Mcdonald": 4, - "Hilke Schellmann": 4, - "IFL Science": 4, - "Isobel Asher Hamilton": 4, - "Jack Morse": 4, - "James Cook": 4, - "Jeff Larson": 4, - "Johana Bhuiyan": 4, - "Jon Brodkin": 4, - "Jordan Pearson": 4, - "Joseph Cox": 4, - "Kevin Roose": 4, - "Khari Johnson": 4, - "Kyle Wiggers": 4, - "Lauren Kirchner": 4, - "Lauren Leffer": 4, - "Lucas Ropek": 4, - "Luke Henriques-Gomes": 4, - "Madison Malone Kircher": 4, - "Maggie Harrison Dupré": 4, - "Matt Novak": 4, - "National Highway Traffic Safety Administration": 4, - "Privacy International": 4, - "RT": 4, - "Rachel Metz": 4, - "Rob Price": 4, - "Sean Keach": 4, - "Shannon Liao": 4, - "Srishti Deoras": 4, - "The Inquirer": 4, - "Thomas Orsolya": 4, - "Wes Davis": 4, - "Aaron Rieke": 3, - "Abigail Beall": 3, - "Adam Satariano": 3, - "Adrienne Roberts": 3, - "Aj Dellinger": 3, - "Al Jazeera": 3, - "Alan Levin": 3, - "Aleksandra Korolova": 3, - "Alex Kasprak": 3, - "Alex Seitz-Wald": 3, - "Ananya Bhattacharya": 3, - "Andrew Higgins": 3, - "Annie Palmer": 3, - "Ariel Zilber": 3, - "Australian Broadcasting Corporation": 3, - "Bill McCarthy": 3, - "Blake Brittain": 3, - "Bobby Allyn": 3, - "Brendan Pierson": 3, - "Brett Molina": 3, - "Brian Barrett": 3, - "Cara McGoogan": 3, - "Carl Franzen": 3, - "Caroline Haskins": 3, - "Catherine Thorbecke": 3, - "Cecily Mauran": 3, - "Charles Pulliam-Moore": 3, - "Chris Burt": 3, - "City News Service": 3, - "Cory Doctorow": 3, - "Dan Levine": 3, - "Daniel Wu": 3, - "Dara Kerr": 3, - "Davey Alba": 3, - "David Gilbert": 3, - "David Grossman": 3, - "David Lumb": 3, - "Department of Motor Vehicles": 3, - "Devin Coldewey": 3, - "Dhruv Mehrotra": 3, - "Douglas Heaven": 3, - "Douglas Macmillan": 3, - "Duncan Riley": 3, - "Eileen Guo": 3, - "Elizabeth Dwoskin": 3, - "Elizabeth Napolitano": 3, - "Emilia David": 3, - "Erielle Delzer": 3, - "Faustine Ngila": 3, - "Foxglove": 3, - "Frank Landymore": 3, - "Gareth Corfield": 3, - "Geoffrey A. Fowler": 3, - "Gianluca Mezzofiore": 3, - "Global Witness": 3, - "Heather Somerville": 3, - "Heather Vogell": 3, - "Hyunjoo Jin": 3, - "IANS": 3, - "Issie Lapowsky": 3, - "Jack Gillum": 3, - "Jackie Snow": 3, - "Jane Wakefield": 3, - "Jeff Horwitz": 3, - "Jefferson Graham": 3, - "Jeffrey Dastin": 3, - "Jim Waterson": 3, - "Jonathan M. Gitlin": 3, - "Kevin Okemwa": 3, - "Kieren McCarthy": 3, - "Kit Eaton": 3, - "Kris Holt": 3, - "Kyle Orland": 3, - "Lauren Kaori Gurley": 3, - "Leon Yin": 3, - "Louise Matsakis": 3, - "Loukia Papadopoulos": 3, - "Luke Dormehl": 3, - "Mallory Locklear": 3, - "Marco Marcelline": 3, - "Matt Binder": 3, - "Matt O'Brien": 3, - "Megan Cerullo": 3, - "Mia Sato": 3, - "Michael Kan": 3, - "Mikael Thalen": 3, - "Mike Murphy": 3, - "Morgan Meaker": 3, - "Naomi Nix": 3, - "National Transportation Safety Board": 3, - "Natt Garun": 3, - "Neil Bedi": 3, - "News Desk": 3, - "Nicholas Thompson": 3, - "Nick Robins-Early": 3, - "Nicolas Kayser-Bril": 3, - "OpenAI": 3, - "Pandaily": 3, - "Paris Martineau": 3, - "Patrick Collinson": 3, - "Paul Mozur": 3, - "Phoebe Weston": 3, - "Phys Org": 3, - "Reuters Editorial": 3, - "Rob Beschizza": 3, - "Rob Stumpf": 3, - "Robert Booth": 3, - "Ronan Glon": 3, - "Ryan Whitwam": 3, - "Sam Biddle": 3, - "Saqib Shah": 3, - "Sejal Sharma": 3, - "Sheera Frenkel": 3, - "Stephen Edelstein": 3, - "Synced": 3, - "Team Latestly": 3, - "The Straits Times": 3, - "Thomas Barrabi": 3, - "Thomas Brewster": 3, - "Thomas Germain": 3, - "Thomas Macaulay": 3, - "Tim De Chant": 3, - "Times of India": 3, - "Tom Warren": 3, - "Trisha Thadani": 3, - "Tristan Greene": 3, - "ABP News Bureau": 2, - "Aaron Glantz": 2, - "Aaron Horowitz": 2, - "Aaron Mok": 2, - "Aaron Sankin": 2, - "Aaron Schaffer": 2, - "Abby Ohlheiser": 2, - "Abhirup Roy": 2, - "Adam Clark Estes": 2, - "Adi Robertson": 2, - "Adrianna Nine": 2, - "Adrianne Jeffries": 2, - "Adrienne Lafrance": 2, - "Aimee Picchi": 2, - "Alan Martin": 2, - "Alan Mislove": 2, - "Alex Davies": 2, - "Alex Pigman": 2, - "Alfred Ng": 2, - "Alisha Rahaman Sarkar": 2, - "Alistair Barr": 2, - "Alistair Charlton": 2, - "Allana Akhtar": 2, - "Alyssa Mercante": 2, - "Amanda Hoover": 2, - "Amnesty International": 2, - "Amrita Khalid": 2, - "Ana Gutiérrez": 2, - "Andrew Griffin": 2, - "Andrew Hobbs": 2, - "Andrew Liptak": 2, - "Andy Greenberg": 2, - "Angela Yang": 2, - "Anjana Samant": 2, - "Anne Hayes": 2, - "Annie Gilbertson": 2, - "Anonymous": 2, - "Anthony Cuthbertson": 2, - "Antoine Allen": 2, - "Aol Staff": 2, - "App Drivers and Couriers Union": 2, - "April Glaser": 2, - "April Rubin": 2, - "Ariana Tobin": 2, - "Ariel Bogle": 2, - "Ariel Herbert-Voss": 2, - "Arijeta Lajka": 2, - "Arvind Narayanan": 2, - "Arwa Mahdawi": 2, - "Avi Asher-Schapiro": 2, - "Ayang Macdonald": 2, - "Barbara Ortutay": 2, - "Beatrice Nolan": 2, - "Bella Cacciatore": 2, - "Ben Dickson": 2, - "Ben Ellery": 2, - "Beth Mole": 2, - "Bill Donahue": 2, - "Billy Perrigo": 2, - "Blake Montgomery": 2, - "Bob Herman": 2, - "Bradford Betz": 2, - "Brian Bushard": 2, - "Brian Fung": 2, - "Buster Hein": 2, - "CNN Business": 2, - "CNN Newsource": 2, - "Cailin Loesch": 2, - "Caitlin Kelly": 2, - "Carlos E. Perez": 2, - "Caroline Mimbs Nyce": 2, - "Caroline O'Donovan": 2, - "Casey Tolan": 2, - "Cathy O'Neil": 2, - "Chaim Gartenberg": 2, - "Charles Duhigg": 2, - "Charlie Warzel": 2, - "Chitra Ramaswamy": 2, - "Choe Sang-Hun": 2, - "Chris Isidore": 2, - "Chris Matyszczyk": 2, - "Chris Morris": 2, - "Chris Stokel-Walker": 2, - "Chris Vallance": 2, - "Christian De Looper": 2, - "Christopher Rosa": 2, - "Chuck Dinerstein": 2, - "Chuong Nguyen": 2, - "Ciaran Lyons": 2, - "Clare Duffy": 2, - "Colin Lecher": 2, - "Conor Cawley": 2, - "Dan Robitzski": 2, - "Daniel Croft": 2, - "Daniel E. Ho": 2, - "Danny Palmer": 2, - "David Koenig": 2, - "David Z. Morris": 2, - "De Elizabeth": 2, - "Debra Cassens Weiss": 2, - "Dell Cameron": 2, - "Department of Justice": 2, - "Donie O'Sullivan": 2, - "Dustin Volz": 2, - "Ed Yong": 2, - "Edward Helmore": 2, - "Elizabeth Culliford": 2, - "Emily Dreyfuss": 2, - "Emily Flitter": 2, - "Emmanuelle Saliba": 2, - "Eric Wallace": 2, - "Erin McCormick": 2, - "Ethan Baron": 2, - "Florian Tramèr": 2, - "France 24": 2, - "Garrett M. Graff": 2, - "Gary Craig": 2, - "Gene Maddaus": 2, - "George Joseph": 2, - "Georgia Wells": 2, - "Giulia Carbonaro": 2, - "Globe Staff": 2, - "Grégoire Sauvage": 2, - "Guardian staff and agencies": 2, - "Gustavo Henrique Ruffo": 2, - "Hamza Shaban": 2, - "Hannah Devlin": 2, - "Hannah Getahun": 2, - "Hannah Knowles": 2, - "Hany Farid": 2, - "Harry Suhartono": 2, - "Heather Chen": 2, - "Heather Stewart": 2, - "Henry Belot": 2, - "Hope Reese": 2, - "Igor Bonifacic": 2, - "India Today Web Desk": 2, - "Indo-Asian News Service": 2, - "Insurance Journal": 2, - "Inverse": 2, - "Isaac Chotiner": 2, - "Isaiah Richard": 2, - "Itech Post": 2, - "Ivan Mehta": 2, - "Jack Brewster": 2, - "Jack Smith Iv": 2, - "Jackie Lieberman": 2, - "Jacob Serebrin": 2, - "Jacob Snow": 2, - "Jake Offenhartz": 2, - "Jake Peterson": 2, - "James Clayton": 2, - "James Felton": 2, - "James Purtill": 2, - "James Zou": 2, - "Janus Rose": 2, - "Jarni Blakkarly": 2, - "Jazper Lu": 2, - "Jean Mackenzie": 2, - "Jeff Kao": 2, - "Jennifer Cowan": 2, - "Jennifer L. Doleac": 2, - "Jennifer Langston": 2, - "Jeremy B. Merrill": 2, - "Jeremy Kahn": 2, - "Jess Weatherbed": 2, - "Jessica Guynn": 2, - "Joan Lowy": 2, - "Jodi Kantor": 2, - "Joe Kukura": 2, - "Joe Patrice": 2, - "Joe Vaccarelli": 2, - "Joel Hruska": 2, - "John Glenday": 2, - "John Pring": 2, - "John Simerman": 2, - "John West": 2, - "Jonathan Limehouse": 2, - "Jonathan Vanian": 2, - "Joseph Hincks": 2, - "Joseph Menn": 2, - "Joshua Bote": 2, - "Joshua Thurston": 2, - "Julia Carrie Wong": 2, - "Julian E. Barnes": 2, - "Julie Johnsson": 2, - "Justin Pritchard": 2, - "Jyoti Mann": 2, - "Kalhan Rosenblatt": 2, - "Karin Kovary Solymos": 2, - "Kate Proctor": 2, - "Kath Xu": 2, - "Katherine Lee": 2, - "Kathleen Magramo": 2, - "Katie Notopoulos": 2, - "Keith Wagstaff": 2, - "Kelvin Chan": 2, - "Keoni Everington": 2, - "Khamila Mulia": 2, - "Kristian Lum": 2, - "Kyle Barr": 2, - "Kylie Cheung": 2, - "Lara Pearce": 2, - "Larry Alton": 2, - "Latanya Sweeney": 2, - "Lauren Rhue": 2, - "Lauren Sforza": 2, - "Leena Nasir": 2, - "Leonardo Nicoletti": 2, - "Lewin Day": 2, - "Li Tao": 2, - "Liam Reilly": 2, - "Lily Hay Newman": 2, - "Linette Lopez": 2, - "Lorenzo Arvanitis": 2, - "Lucas Manfredi": 2, - "Lydia Horne": 2, - "Maia Szalavitz": 2, - "Marco Della Cava": 2, - "Marie Solis": 2, - "Mariella Moon": 2, - "Marissa Gerchick": 2, - "Mark Frauenfelder": 2, - "Mark Sweney": 2, - "Mary Papenfuss": 2, - "Matt Burgess": 2, - "Matt Simon": 2, - "Matt Wille": 2, - "Matthew Butterick": 2, - "Matthew Guariglia": 2, - "Matthew Humphries": 2, - "Matthew Jagielski": 2, - "Matthew Phelan": 2, - "Matthias Bastian": 2, - "Maya Oppenheim": 2, - "McKenzie Sadeghi": 2, - "Megan Farokhmanesh": 2, - "Megan Hickey": 2, - "Melia Robinson": 2, - "Melissa del Bosque": 2, - "Michael L. Diamond": 2, - "Michael Savage": 2, - "Michael Tarm": 2, - "Michelle Butterfield": 2, - "Mick Akers": 2, - "Microsoft": 2, - "Mike Masnick": 2, - "Mike Melanson": 2, - "Mike Spector": 2, - "Mirna Alsharif": 2, - "Mitchell Clark": 2, - "Moinak Pal": 2, - "Molly Liebergall": 2, - "Moohita Kaur Garg": 2, - "Morgan Sung": 2, - "Muhammad Ali": 2, - "Nadine Freischlad": 2, - "Nam Hyun-woo": 2, - "Natalie O'Neill": 2, - "Natasha Duarte": 2, - "Nathaniel Gleicher": 2, - "Neil Vigdor": 2, - "Niamh Ancell": 2, - "Nicholas Diakopoulos": 2, - "Nicholas Lezard": 2, - "Nick Evershed": 2, - "Nick Statt": 2, - "Nicole Clark": 2, - "Nikki Williams": 2, - "Nilay Patel": 2, - "Nilesh Christopher": 2, - "Niniek Karmini": 2, - "Noam Shemtov": 2, - "Onur Demirkol": 2, - "Paige Skinner": 2, - "Paul Egan": 2, - "Paul Lukas": 2, - "Pete Bigelow": 2, - "Philip Marcelo": 2, - "Phillip Tracy": 2, - "Piotr Sapiezynski": 2, - "Pocharapon Neammanee": 2, - "Political Reporter": 2, - "RT Staff": 2, - "Rachel Kraus": 2, - "Rachel Lerman": 2, - "Ramishah Maruf": 2, - "Rasha Ali": 2, - "Raymond Wong": 2, - "Rebecca Hill": 2, - "Rene Marsh": 2, - "Rita Liao": 2, - "Rob Thubron": 2, - "Rob Waugh": 2, - "Romy Ellenbogen": 2, - "Russ Mitchell": 2, - "Ryan Daws": 2, - "Ryan McNeal": 2, - "Sally Ho": 2, - "Sapna Maheshwari": 2, - "Sara Ashley O'Brien": 2, - "Sarah Frier": 2, - "Sarah Perez": 2, - "Sayash Kapoor": 2, - "Scott Nover": 2, - "Sean B. McGregor": 2, - "Sean McGregor": 2, - "Sean Szymkowski": 2, - "Sebastien Bell": 2, - "Shannon Thaler": 2, - "Shanti Das": 2, - "Sharon Adarlo": 2, - "Sharon Goldman": 2, - "Shira Ovide": 2, - "Shivali Best": 2, - "Shona Ghosh": 2, - "Sigal Samuel": 2, - "Siladitya Ray": 2, - "Simon Willison": 2, - "Soo Youn": 2, - "Sophia Waterfield": 2, - "Sophie Beiers": 2, - "Sophie Nieto-Munoz": 2, - "Staff": 2, - "Stan Schroeder": 2, - "Steph Maj Swanson": 2, - "Stephanie Wykstra": 2, - "Stephen Kafeero": 2, - "Steve Crowe": 2, - "Steve Stecklow": 2, - "Steven Loveday": 2, - "Steven Musil": 2, - "Stuff": 2, - "Tarak Shah": 2, - "Tate Ryan-Mosley": 2, - "Tatum Hunter": 2, - "Taylor Donovan Barnett": 2, - "Tech Transparency Project": 2, - "The Express Tribune": 2, - "The Hindu Bureau": 2, - "The Independent": 2, - "Tim McNicholas": 2, - "Tobi Jegede": 2, - "Tony Ho Tran": 2, - "Tony Kovaleski": 2, - "Tony Tran": 2, - "Turkish Minute": 2, - "Varsha Bansal": 2, - "Victoria Woollaston": 2, - "Vittoria Elliott": 2, - "Wall Street Journal": 2, - "Will Bedingfield": 2, - "Will McCurdy": 2, - "William Gavin": 2, - "William Hoffman": 2, - "William Isaac": 2, - "Yiwen Lu": 2, - "Yuval Abraham": 2, - "bne IntelliNews": 2, - "천호성": 2, - "@BeauHD": 1, - "@DearKick": 1, - "@FSD_in_6m": 1, - "@Robert_AIZI": 1, - "@Waqas": 1, - "@an_open_mind": 1, - "@kawazacky": 1, - "@msravi": 1, - "@ohgustie": 1, - "A.J. Perez": 1, - "AAP": 1, - "ABC 7 Eyewitness News": 1, - "ABC Australia": 1, - "AI Addict": 1, - "AP/Reuters": 1, - "Aaditi Lele": 1, - "Aakar Patel": 1, - "Aamir Sheikh": 1, - "Aaron Blake": 1, - "Aaron Gordon": 1, - "Aaron Gregg": 1, - "Aaron Holmes": 1, - "Aaron Huff": 1, - "Aaron Katersky": 1, - "Aaron Krolik": 1, - "Aaron Mak": 1, - "Aaron Mamiit": 1, - "Aaron Moss": 1, - "Aaron Smith": 1, - "Aaron Tilley": 1, - "Aashish Kumar Shrivastava": 1, - "Aatif Sulleyman": 1, - "Abbie Richards": 1, - "Abby Monteil": 1, - "Abdul Moeed": 1, - "Abdullah": 1, - "Abeba Birhane": 1, - "Abigail Rubenstein": 1, - "Abrar Al-Heeti": 1, - "Abubakar Abid": 1, - "Abubakar Idris": 1, - "Adam Carlson": 1, - "Adam Downer": 1, - "Adam Garfinkle": 1, - "Adam Hadhazy": 1, - "Adam Healy": 1, - "Adam Kalai": 1, - "Adam Lam": 1, - "Adam Rawnsley": 1, - "Adam Roberts": 1, - "Adam Rowe": 1, - "Aditi Sen": 1, - "Aditya Bahl": 1, - "Aditya Kalra": 1, - "Adrian Hopgood": 1, - "Adrian Morrow": 1, - "Adrian Weckler": 1, - "Adriana Lee": 1, - "Agence France Presse": 1, - "Aicha El Hammar Castano": 1, - "Aiden Pink": 1, - "Aimee Lucido": 1, - "Aimee Ortiz": 1, - "Aisha Malik": 1, - "Akash Pandey": 1, - "Aki Anastasiou": 1, - "Aki Peritz": 1, - "Al-Monitor Staff": 1, - "Alain Sherter": 1, - "Alan Boyle": 1, - "Alan Friedman": 1, - "Alastair Gale": 1, - "Albert Tait": 1, - "Alberto Luperon": 1, - "Alberto Romero": 1, - "Aleks Phillips": 1, - "Alessandra Kellermann": 1, - "Alessandro Mascellino": 1, - "Alex Arger": 1, - "Alex Brenninkmeijer": 1, - "Alex Cadier": 1, - "Alex Carr": 1, - "Alex Clark": 1, - "Alex Farber": 1, - "Alex Heath": 1, - "Alex Isenstadt": 1, - "Alex Johnson": 1, - "Alex Kaplan": 1, - "Alex Kierstein": 1, - "Alex Lockie": 1, - "Alex Mitchell": 1, - "Alex Moersen": 1, - "Alex Portée": 1, - "Alex Tsiaoussidis": 1, - "Alex Wawro": 1, - "Alexa Cimino": 1, - "Alexa Corse": 1, - "Alexa Hagerty": 1, - "Alexander Hanff": 1, - "Alexander J Martin": 1, - "Alexander Martin": 1, - "Alexander Puutio": 1, - "Alexandra Blogier": 1, - "Alexandra Kravariti": 1, - "Alexandra Marquez": 1, - "Alexandra Miller": 1, - "Alexandra S. Levine": 1, - "Alexei Alexis": 1, - "Alexis Haden": 1, - "Alexis Kleinman": 1, - "Alexis Madrigal": 1, - "Ali A. Jessani": 1, - "Ali Breland": 1, - "Ali Swenson": 1, - "Alice Hearing": 1, - "Alice Kohn": 1, - "Alice Milliken": 1, - "Alice Richardson": 1, - "Alice Wright": 1, - "Alicia Valenski": 1, - "Alina Oprea": 1, - "Alisha Green": 1, - "Alison Cutler": 1, - "Alison Sider": 1, - "Alix Langone": 1, - "Alizeh Kohari": 1, - "Allison Koenecke": 1, - "Allison Levitsky": 1, - "Alma Fabiani": 1, - "Alvin E. Roth": 1, - "Alyce McFadden": 1, - "Alyse Stanley": 1, - "Alyssa Braithwaite": 1, - "Alyssa Goard": 1, - "Alyssa Guzman": 1, - "Alyssa Newcomb": 1, - "Alyssa Rinelli": 1, - "Amal Attar-Guzman": 1, - "Amanda Claypool": 1, - "Amanda Cochran": 1, - "Amanda Geffner": 1, - "Amanda Kooser": 1, - "Amanda Lewis": 1, - "Amanda Schupak": 1, - "Amanda Shaw": 1, - "Amaya Ross": 1, - "Amelia Butterly": 1, - "Amelia Gentleman": 1, - "Amelia Lucas": 1, - "Amelia McGuire": 1, - "Amelia Mcdonell-Parry": 1, - "Amiah Taylor": 1, - "Amit Katwala": 1, - "Amiya Moretta": 1, - "Amrutha Pagad": 1, - "Amy Beth Hanson": 1, - "Amy Gardner": 1, - "Amy Kraft": 1, - "Amy Martinez": 1, - "Ana María Arévalo Gosen": 1, - "Anagha Srikanth": 1, - "Anand Tamboli": 1, - "Ananya Gairola": 1, - "Andrea Bernstein": 1, - "Andrea Blanco": 1, - "Andreo Calonzo": 1, - "Andrew Buncombe": 1, - "Andrew Court": 1, - "Andrew Deck": 1, - "Andrew Desiderio": 1, - "Andrew Guthrie Ferguson": 1, - "Andrew Hutchinson": 1, - "Andrew Longhi": 1, - "Andrew Nam": 1, - "Andrew Orlowski": 1, - "Andrew Papachristos": 1, - "Andrew Schlaikjer": 1, - "Andrew Smith": 1, - "Andrew Tangel": 1, - "Andrew Thompson": 1, - "Andrew Trotman": 1, - "Andrew Van Dam": 1, - "Andrew Wilks": 1, - "Andy Brown": 1, - "Andy Miller": 1, - "Andy Mukherjee": 1, - "Andy Nghiem": 1, - "Andy Pasztor": 1, - "Andy Verity": 1, - "Angel Saunders": 1, - "Angela B.": 1, - "Angela Lai": 1, - "Angelica Mari": 1, - "Angry Asian Man": 1, - "Angus Morrison": 1, - "Ani": 1, - "Ani Nenkova": 1, - "Anisa Harrasy": 1, - "Anjana Nair": 1, - "Ankita Garg": 1, - "Anna Andrews": 1, - "Anna Bahney": 1, - "Anna Houlahan": 1, - "Anna Iovine": 1, - "Anna Kutz": 1, - "Anna Loup": 1, - "Anna Lucente Sterling": 1, - "Anna Tong": 1, - "Annalyn Kurtz": 1, - "Anne Marie Lee": 1, - "Anneli L. Tostar": 1, - "Annie Minoff": 1, - "Annie Palmer For Dailymail.Com": 1, - "Anthony G. Attrino": 1, - "Antoine Gara": 1, - "Anton Ekker": 1, - "Antonio Andrés-Pueyo": 1, - "Antonio Madeira": 1, - "Antonio Pequeño IV": 1, - "Antonio Villas-Boas": 1, - "Anubhav Sharma": 1, - "Anumita Kaur": 1, - "Anupam Dikhit": 1, - "Anupama Airy": 1, - "Anurag Baruah": 1, - "Anushe Fawaz": 1, - "Anushree Hede": 1, - "Aparna Iyer": 1, - "Apple On Amazon": 1, - "Archie Hamilton": 1, - "Archie Mitchell": 1, - "Ardi Wirdana": 1, - "Arfa Javaid": 1, - "Ariana Baio": 1, - "Arianna Evers": 1, - "Aric Jenkins": 1, - "Arielle Pardes": 1, - "Arif Perdana": 1, - "Aris Folley": 1, - "Arjun Kharpal": 1, - "Armin Ronacher": 1, - "Armin Rosen": 1, - "Artemis Moshtaghian": 1, - "Arthur Holland Michel": 1, - "Arun Ramesh": 1, - "Aryan Prakash": 1, - "Aryn Plax": 1, - "Ashleigh Hollowell": 1, - "Ashley Bardhan": 1, - "Ashley Capoot": 1, - "Ashley Collman": 1, - "Ashley Halsey III": 1, - "Ashley Hume": 1, - "Ashley May": 1, - "Ashley McBride": 1, - "Ashley Rodriguez": 1, - "Ashlie D. Stevens": 1, - "Ashok Ramprasad": 1, - "Ashutosh Tripathi": 1, - "Asia Grace": 1, - "Aspyhackr": 1, - "Atlanta Journal-Constitution": 1, - "Audie Cornish": 1, - "Audrey Ash": 1, - "Augustine Fou": 1, - "Austin Mullen": 1, - "Australian Associated Press": 1, - "Ava Mutchler": 1, - "Avaaz": 1, - "Averi Kremposky": 1, - "Avinash A": 1, - "Aye Min Thant": 1, - "Aylin Caliskan": 1, - "Aylin Elci": 1, - "Ayushman Kaul": 1, - "B J Robertson": 1, - "BR Data": 1, - "Ban Barkawi": 1, - "Barbara Liston": 1, - "Barbara S. Peterson": 1, - "Barbara Zmušková": 1, - "Barry Sookman": 1, - "Basileal Imana": 1, - "Bayu Anggorojati": 1, - "Bbc Trending": 1, - "Beatrice Verhoeven": 1, - "Becca Smouse": 1, - "Becky Bargh": 1, - "Beebom": 1, - "Beijing Newsroom": 1, - "Bel Air Association": 1, - "Belga News Agency": 1, - "Belinda Grant Geary": 1, - "Ben Beaumont-Thomas": 1, - "Ben Brumfield": 1, - "Ben Chu": 1, - "Ben Cohen": 1, - "Ben Cost": 1, - "Ben Dubow": 1, - "Ben Eltham": 1, - "Ben Geman": 1, - "Ben Guarino": 1, - "Ben Kaufman": 1, - "Ben Lovejoy": 1, - "Ben Nimmo": 1, - "Ben Popper": 1, - "Ben Sales": 1, - "Ben Strauss": 1, - "Ben Sullivan": 1, - "Ben Travis": 1, - "Ben Warwick": 1, - "Ben Woods": 1, - "Ben Yelin": 1, - "Benedict Smith": 1, - "Benita Kolovos": 1, - "Benjamin Fearnow": 1, - "Benjamin Goggin": 1, - "Benjamin Haas": 1, - "Benjamin Lindsay": 1, - "Benjamin Perkin": 1, - "Benjamin Schneider": 1, - "Benjamin Snyder": 1, - "Benjamin Wallace-Wells": 1, - "Berenice Baker": 1, - "Berhan Taye": 1, - "Bernie Woodall": 1, - "Beth Greenfield": 1, - "Bethan McKernan": 1, - "Bethan Sexton": 1, - "Bethany Hallam": 1, - "Bethany White": 1, - "Bettina Büchel": 1, - "Bev Stephans": 1, - "Bharat Sharma": 1, - "Bhavya Sukheja": 1, - "Bianca Bosker": 1, - "Bianca Britton": 1, - "Bigger Law Firm Magazine": 1, - "Bijan Stephen": 1, - "Bill Cash": 1, - "Bill Christensen": 1, - "Bill Snyder": 1, - "Biman Mukherji": 1, - "Bindu Bansinath": 1, - "Biometrics Commissioner": 1, - "Björn ten Seldam": 1, - "Black Voices Editor": 1, - "Blair Hanley Frank": 1, - "Blake Harper": 1, - "Blake Lew-Merwin": 1, - "Bobbie Johnson": 1, - "Bogdan Popa": 1, - "Boston Herald": 1, - "Botego Inc": 1, - "Brad Anderson": 1, - "Braden Bjella": 1, - "Bradley Hope": 1, - "Bradley Jolly": 1, - "Brandon Turkus": 1, - "Brayden Lindrea": 1, - "Breanna Barraclough": 1, - "Bree Fowler": 1, - "Breeze Liu": 1, - "Brenda Medina": 1, - "Brenden Gallagher": 1, - "Brett Wilkins": 1, - "Brian Fraga": 1, - "Brian Lisi": 1, - "Brian Melley": 1, - "Brian Merchant": 1, - "Brian Resnick": 1, - "Brian Welk": 1, - "Brian Whitton": 1, - "Brianna Sacks": 1, - "Bridget Chavez": 1, - "Bridget McArthur": 1, - "Brie Barbee": 1, - "Brie Stimson": 1, - "Brieanna J Frank": 1, - "Brinda A. Thomas": 1, - "Brittny Mejia": 1, - "Bruce Brown": 1, - "Bruce Handy": 1, - "Bruce Schneier": 1, - "Bryan Clark": 1, - "Bryan Logan": 1, - "Bálint Miklós": 1 - }, - "language": { - "en": 3838, - "de": 6, - "es": 5, - "fr": 4, - "it": 4, - "th": 4, - "vi": 4, - "ko": 2, - "hi": 1, - "nl": 1, - "pt": 1, - "zh-CN": 1 - }, - "namespaces": { - "CSETv1_Annotator-1": 1898, - "CSETv1": 1617, - "CSETv0": 1217, - "CSETv1_Annotator-2": 954, - "CSETv1_Annotator-3": 436, - "GMF": 402 - }, - "submitters": { - "Daniel Atherton": 1349, - "Anonymous": 862, - "Khoa Lam": 416, - "Roman Yampolskiy": 372, - "Catherine Olsson": 208, - "Kate Perkins": 139, - "Ingrid Dickinson (CSET)": 73, - "Thomas Giallella (CSET)": 68, - "Sean McGregor": 42, - "Patrick Hall": 33, - "Srishti Khemka (CSET)": 33, - "Roman Lutz": 26, - "Devon Colmer (CSET)": 22, - "CSET annotators": 21, - "Sonali Pednekar (CSET)": 21, - "Logan B": 20, - "Collin Starkweather": 15, - "Janet Schwartz": 14, - "Luna McNulty": 11, - "Kevin Paeth": 10, - "Nickie Demakos": 9, - "Andrew Hundt": 7, - "Frank Guo": 7, - "Ayrton San Joaquin": 6, - "Neama Dadkhahnikoo": 6, - "Arthit Suriyawongkul": 5, - "Andrew Gamino-Cheong": 4, - "Michael Simon (ForHumanity)": 4, - "Viviana Rivera": 4, - "Catharina Doria": 3, - "Cesar Varela": 3, - "Helen Zhu": 3, - "Sundar Narayanan (ForHumanity)": 3, - "Alice RT (ForHumanity)": 2, - "Charlie Pownall": 2, - "Effy Elden": 2, - "Fionntan O'Donnell": 2, - "Irina Borisova King": 2, - "Maud Stiernet (ForHumanity)": 2, - "Nathan Butters": 2, - "Nik Martelaro": 2, - "Octavia Occident": 2, - "Sundar Narayanan": 2, - "William Schindhelm Georg": 2, - "kepa": 2, - "reubot": 2, - "21five": 1, - "Alexis Monks": 1, - "Alice Villano": 1, - "Antonio Buffelli (ForHumanity)": 1, - "Ashley Casovan": 1, - "Beatrice Moissinac": 1, - "Carol Anderson (ForHumanity)": 1, - "Charlie Wang": 1, - "Christopher Maratos": 1, - "Corey Abshire": 1, - "Eric Horvitz": 1, - "Fabio Xie": 1, - "Fion Lee-Madan (Fairly AI)": 1, - "Gerry Chng": 1, - "Inbal - For Humanity": 1, - "Inbal - ForHumanity": 1, - "Isabelle": 1, - "Joanna (ForHumanity)": 1, - "Jodi Masters-Gonzales": 1, - "Johann Hanssen (ForHumanity)": 1, - "Joshua Poore": 1, - "K M": 1, - "Karson Elmgren": 1, - "Leon Overweel": 1, - "Lilianna Smith": 1, - "Lilly Ryan": 1, - "Madison Malone Kircher": 1, - "Matteo Dora": 1, - "Mitt Regan Jr. (BABL AI)": 1, - "Natalia DLC": 1, - "Nga Than": 1, - "Nick Stockton": 1, - "Parul Pandey": 1, - "Ryan Carrier (ForHumanity)": 1, - "Samuel Curtis": 1, - "Scott Cambo": 1, - "Subhabrata Majumdar": 1, - "Subho Majumdar": 1, - "TNT": 1, - "Tuoi Tran": 1, - "Wiebke Hutiri": 1, - "Yukti Handa": 1, - "kepae": 1 - }, - "incident_id": { - "1": 14, - "2": 17, - "3": 19, - "4": 25, - "5": 12, - "6": 28, - "7": 6, - "8": 10, - "9": 7, - "10": 10, - "11": 15, - "12": 1, - "13": 9, - "14": 7, - "15": 24, - "16": 24, - "17": 22, - "18": 11, - "19": 27, - "20": 22, - "22": 22, - "23": 24, - "24": 27, - "25": 11, - "26": 24, - "27": 27, - "28": 30, - "29": 2, - "30": 28, - "31": 29, - "32": 21, - "33": 4, - "34": 35, - "35": 20, - "36": 25, - "37": 33, - "38": 11, - "39": 29, - "40": 22, - "41": 28, - "42": 1, - "43": 4, - "44": 1, - "45": 29, - "46": 6, - "47": 9, - "48": 22, - "49": 10, - "50": 24, - "51": 27, - "52": 29, - "53": 18, - "54": 13, - "55": 16, - "56": 7, - "57": 39, - "58": 5, - "59": 10, - "60": 23, - "61": 1, - "63": 1, - "64": 1, - "65": 1, - "66": 16, - "67": 24, - "68": 30, - "69": 12, - "70": 4, - "71": 28, - "72": 26, - "73": 8, - "74": 11, - "75": 1, - "76": 1, - "77": 5, - "78": 1, - "79": 3, - "80": 2, - "81": 1, - "82": 1, - "83": 1, - "84": 1, - "86": 2, - "87": 1, - "88": 2, - "89": 1, - "91": 5, - "92": 6, - "93": 4, - "94": 2, - "95": 4, - "96": 1, - "97": 1, - "98": 1, - "99": 1, - "100": 1, - "101": 6, - "102": 2, - "103": 5, - "104": 1, - "105": 1, - "106": 13, - "107": 2, - "108": 3, - "109": 1, - "110": 1, - "111": 5, - "112": 11, - "113": 1, - "114": 1, - "115": 3, - "116": 2, - "117": 4, - "118": 3, - "119": 4, - "120": 1, - "121": 4, - "122": 1, - "123": 4, - "124": 7, - "125": 3, - "126": 4, - "127": 12, - "128": 2, - "129": 1, - "131": 2, - "132": 1, - "133": 1, - "134": 2, - "135": 2, - "136": 1, - "137": 1, - "138": 6, - "139": 2, - "140": 1, - "141": 2, - "142": 1, - "143": 1, - "144": 6, - "145": 3, - "146": 3, - "147": 2, - "148": 1, - "149": 4, - "150": 3, - "151": 10, - "152": 2, - "153": 3, - "154": 1, - "155": 2, - "156": 2, - "157": 1, - "158": 1, - "160": 2, - "161": 3, - "162": 1, - "163": 2, - "164": 1, - "165": 2, - "166": 2, - "167": 1, - "168": 2, - "169": 5, - "170": 3, - "171": 1, - "172": 3, - "173": 1, - "174": 4, - "175": 5, - "176": 1, - "177": 5, - "178": 8, - "179": 3, - "180": 3, - "181": 2, - "182": 2, - "183": 6, - "184": 3, - "185": 4, - "186": 7, - "187": 3, - "188": 4, - "189": 6, - "190": 4, - "191": 2, - "192": 2, - "193": 1, - "194": 1, - "195": 12, - "196": 5, - "197": 4, - "198": 9, - "199": 7, - "200": 1, - "201": 2, - "202": 5, - "203": 3, - "204": 4, - "205": 4, - "206": 4, - "207": 4, - "208": 8, - "209": 5, - "210": 3, - "211": 7, - "212": 4, - "213": 5, - "214": 1, - "215": 1, - "216": 5, - "217": 2, - "218": 3, - "219": 1, - "220": 4, - "221": 2, - "222": 1, - "223": 1, - "224": 1, - "225": 2, - "226": 3, - "227": 3, - "228": 1, - "229": 2, - "230": 1, - "231": 4, - "232": 4, - "233": 1, - "234": 2, - "235": 1, - "236": 1, - "238": 1, - "239": 1, - "240": 5, - "241": 5, - "242": 1, - "243": 2, - "244": 1, - "245": 1, - "246": 2, - "248": 2, - "249": 2, - "250": 1, - "251": 3, - "252": 1, - "253": 4, - "254": 2, - "255": 9, - "256": 1, - "257": 4, - "258": 2, - "259": 2, - "260": 2, - "261": 8, - "262": 4, - "263": 1, - "264": 1, - "265": 2, - "266": 8, - "267": 10, - "268": 4, - "270": 1, - "271": 3, - "272": 3, - "273": 1, - "274": 2, - "275": 2, - "276": 1, - "277": 1, - "278": 3, - "279": 3, - "280": 2, - "281": 3, - "282": 3, - "283": 1, - "284": 6, - "285": 1, - "286": 2, - "288": 4, - "289": 2, - "290": 3, - "291": 6, - "292": 3, - "293": 7, - "294": 4, - "295": 5, - "296": 3, - "297": 3, - "298": 1, - "299": 1, - "300": 2, - "301": 1, - "302": 1, - "303": 2, - "304": 1, - "305": 2, - "306": 3, - "307": 1, - "308": 2, - "309": 4, - "310": 8, - "311": 2, - "312": 4, - "313": 2, - "314": 1, - "315": 1, - "316": 1, - "317": 1, - "318": 2, - "319": 5, - "320": 4, - "321": 14, - "322": 2, - "323": 4, - "324": 6, - "325": 2, - "326": 3, - "327": 1, - "328": 2, - "329": 1, - "330": 1, - "331": 2, - "332": 4, - "333": 6, - "334": 2, - "335": 8, - "336": 4, - "337": 9, - "339": 14, - "340": 1, - "341": 5, - "343": 2, - "344": 1, - "345": 1, - "346": 5, - "347": 3, - "348": 3, - "349": 2, - "350": 2, - "351": 1, - "352": 4, - "353": 4, - "354": 5, - "355": 4, - "356": 2, - "357": 3, - "358": 1, - "359": 1, - "360": 3, - "361": 1, - "362": 1, - "363": 1, - "364": 1, - "366": 1, - "367": 1, - "368": 10, - "369": 1, - "370": 1, - "371": 3, - "372": 3, - "373": 12, - "374": 8, - "375": 3, - "376": 5, - "377": 1, - "378": 2, - "379": 2, - "380": 5, - "381": 1, - "382": 1, - "383": 2, - "384": 2, - "385": 6, - "386": 3, - "387": 1, - "388": 1, - "389": 2, - "390": 1, - "391": 2, - "392": 2, - "393": 1, - "394": 2, - "395": 4, - "396": 1, - "397": 2, - "398": 3, - "399": 3, - "400": 1, - "401": 4, - "402": 1, - "403": 2, - "404": 2, - "405": 2, - "406": 1, - "407": 1, - "408": 1, - "409": 3, - "410": 1, - "411": 1, - "412": 4, - "413": 2, - "414": 1, - "415": 5, - "416": 3, - "417": 4, - "418": 3, - "419": 3, - "420": 11, - "421": 9, - "422": 1, - "423": 5, - "424": 4, - "425": 2, - "426": 1, - "427": 3, - "428": 3, - "429": 4, - "430": 21, - "431": 3, - "432": 1, - "433": 10, - "434": 8, - "435": 4, - "436": 9, - "437": 4, - "438": 3, - "439": 3, - "440": 6, - "441": 5, - "443": 25, - "444": 1, - "445": 4, - "446": 5, - "447": 2, - "448": 1, - "449": 4, - "450": 8, - "451": 5, - "452": 2, - "453": 1, - "454": 2, - "455": 7, - "456": 7, - "457": 3, - "458": 1, - "459": 2, - "460": 2, - "461": 4, - "462": 5, - "463": 3, - "464": 4, - "465": 1, - "466": 7, - "467": 14, - "468": 4, - "469": 3, - "470": 2, - "471": 8, - "472": 1, - "473": 1, - "474": 1, - "475": 5, - "476": 3, - "477": 5, - "478": 13, - "479": 4, - "480": 15, - "481": 6, - "482": 20, - "483": 1, - "484": 4, - "485": 1, - "486": 5, - "487": 3, - "488": 1, - "489": 1, - "490": 3, - "491": 1, - "492": 7, - "493": 1, - "494": 5, - "495": 2, - "496": 2, - "497": 2, - "498": 2, - "499": 11, - "500": 1, - "501": 1, - "502": 3, - "503": 7, - "504": 1, - "505": 7, - "506": 2, - "507": 2, - "508": 4, - "509": 2, - "510": 5, - "511": 2, - "513": 5, - "514": 1, - "515": 2, - "516": 2, - "517": 2, - "518": 1, - "519": 1, - "520": 2, - "521": 1, - "522": 1, - "523": 1, - "524": 1, - "525": 3, - "526": 2, - "527": 2, - "528": 2, - "529": 3, - "530": 2, - "531": 1, - "532": 1, - "533": 2, - "534": 2, - "535": 2, - "536": 2, - "537": 2, - "538": 5, - "539": 1, - "540": 5, - "541": 58, - "543": 17, - "544": 22, - "545": 46, - "546": 3, - "547": 2, - "548": 1, - "549": 1, - "550": 2, - "551": 2, - "552": 1, - "553": 2, - "554": 1, - "555": 1, - "556": 4, - "557": 4, - "558": 3, - "559": 2, - "560": 1, - "561": 3, - "562": 1, - "563": 2, - "564": 1, - "565": 4, - "566": 1, - "567": 1, - "568": 2, - "569": 2, - "570": 1, - "571": 1, - "572": 1, - "573": 11, - "574": 1, - "575": 1, - "576": 1, - "577": 1, - "578": 1, - "579": 1, - "580": 1, - "581": 1, - "582": 1, - "583": 1, - "584": 1, - "585": 3, - "586": 1, - "587": 1, - "588": 1, - "589": 1, - "590": 1, - "591": 2, - "592": 3, - "593": 1, - "594": 1, - "595": 1, - "596": 1, - "597": 44, - "598": 1, - "599": 2, - "600": 1, - "601": 10, - "602": 7, - "603": 1, - "604": 7, - "605": 10, - "606": 15, - "607": 1, - "608": 21, - "609": 2, - "610": 10, - "611": 13, - "612": 17, - "613": 1, - "614": 1, - "615": 4, - "616": 43, - "617": 2, - "618": 3, - "619": 14, - "620": 5, - "621": 3, - "622": 6, - "623": 12, - "624": 18, - "625": 5, - "626": 30, - "627": 8, - "628": 14, - "629": 2, - "630": 3, - "631": 2, - "632": 31, - "633": 17, - "634": 21, - "635": 1, - "636": 5, - "638": 17, - "639": 4, - "640": 3, - "641": 12, - "642": 5, - "643": 6, - "644": 6, - "645": 35, - "646": 1, - "647": 2, - "648": 2, - "649": 1, - "650": 1, - "651": 2, - "652": 1, - "653": 1, - "654": 1, - "655": 1, - "656": 3, - "657": 1, - "658": 3, - "659": 3, - "660": 1, - "661": 1, - "662": 2, - "663": 1, - "664": 1, - "665": 1, - "666": 1, - "667": 1, - "668": 1, - "669": 1, - "670": 2, - "671": 1, - "672": 7, - "673": 1, - "674": 1, - "675": 4, - "676": 2, - "677": 1, - "678": 1, - "679": 1, - "680": 2, - "681": 1, - "682": 1, - "683": 2, - "684": 1, - "685": 1, - "686": 2, - "687": 1, - "688": 14, - "689": 4, - "690": 4, - "691": 2, - "692": 2, - "693": 7, - "694": 1, - "695": 1, - "696": 1, - "697": 1, - "698": 1, - "699": 2, - "700": 2, - "701": 1, - "702": 1, - "703": 1, - "704": 2, - "705": 2, - "706": 1, - "707": 1, - "708": 1, - "709": 1, - "710": 1, - "711": 2, - "712": 2, - "713": 1, - "714": 2, - "715": 1, - "716": 1, - "717": 1, - "718": 1, - "719": 1, - "720": 1, - "721": 1, - "722": 1, - "723": 2, - "724": 1, - "725": 1, - "726": 4, - "727": 1, - "728": 1, - "729": 1, - "730": 1, - "731": 1, - "732": 1, - "733": 2, - "734": 1, - "735": 1, - "736": 2, - "737": 2, - "738": 5, - "739": 2, - "740": 1, - "741": 1, - "742": 1, - "743": 1, - "744": 1, - "745": 2, - "746": 2, - "747": 1, - "748": 1, - "749": 1, - "750": 1, - "751": 2, - "752": 1, - "753": 1, - "754": 1, - "755": 2, - "756": 3, - "757": 1, - "758": 1, - "759": 1, - "760": 2, - "761": 1, - "762": 3, - "763": 1, - "764": 4, - "765": 5, - "766": 2, - "767": 2, - "768": 1, - "769": 2, - "770": 3, - "771": 1, - "772": 2, - "773": 1, - "774": 2, - "775": 1, - "776": 1, - "777": 7, - "778": 2, - "779": 1, - "780": 1, - "781": 1, - "782": 1, - "783": 1, - "784": 1, - "785": 1, - "786": 2, - "787": 2, - "788": 1, - "789": 1, - "790": 2, - "791": 1, - "792": 10, - "793": 2, - "794": 2, - "795": 1, - "796": 3, - "797": 5, - "798": 1, - "799": 2, - "800": 1, - "801": 1, - "802": 1, - "803": 1, - "804": 5, - "805": 10, - "806": 1, - "807": 7, - "808": 1, - "809": 1, - "810": 1, - "811": 4, - "812": 5, - "813": 2, - "814": 9, - "815": 1, - "816": 1, - "817": 8, - "818": 2, - "819": 1, - "820": 5, - "821": 1, - "822": 2, - "823": 3, - "824": 11, - "825": 2, - "826": 34, - "827": 1, - "828": 3, - "829": 1, - "830": 1, - "831": 1, - "832": 2, - "833": 2, - "834": 2, - "835": 1, - "836": 1, - "837": 1, - "838": 1, - "839": 20, - "840": 1, - "841": 3, - "842": 16, - "843": 2, - "844": 3, - "845": 1, - "846": 1, - "847": 1, - "848": 1, - "849": 1, - "850": 1 - }, - "source_domain": { - "theguardian.com": 148, - "nytimes.com": 124, - "theverge.com": 103, - "washingtonpost.com": 94, - "wired.com": 66, - "bbc.com": 58, - "reuters.com": 54, - "arstechnica.com": 52, - "futurism.com": 52, - "vice.com": 52, - "businessinsider.com": 46, - "forbes.com": 40, - "apnews.com": 39, - "dailymail.co.uk": 39, - "gizmodo.com": 36, - "wsj.com": 36, - "telegraph.co.uk": 33, - "cnn.com": 32, - "cbsnews.com": 31, - "independent.co.uk": 31, - "mashable.com": 30, - "nbcnews.com": 30, - "twitter.com": 30, - "theregister.com": 29, - "fortune.com": 28, - "nypost.com": 28, - "qz.com": 28, - "engadget.com": 26, - "abc.net.au": 25, - "technologyreview.com": 25, - "cnbc.com": 24, - "techcrunch.com": 23, - "usatoday.com": 22, - "cnet.com": 21, - "medium.com": 19, - "npr.org": 19, - "theregister.co.uk": 15, - "arxiv.org": 14, - "bloomberg.com": 14, - "buzzfeednews.com": 14, - "foxnews.com": 14, - "thetimes.co.uk": 14, - "time.com": 14, - "digitaltrends.com": 13, - "dailydot.com": 12, - "edition.cnn.com": 12, - "electrek.co": 12, - "theatlantic.com": 12, - "thehill.com": 12, - "venturebeat.com": 12, - "abcnews.go.com": 11, - "businessinsider.com.au": 11, - "latimes.com": 11, - "news.sky.com": 11, - "slate.com": 11, - "thenextweb.com": 11, - "yahoo.com": 11, - "gizmodo.com.au": 10, - "motherboard.vice.com": 10, - "pcmag.com": 10, - "propublica.org": 10, - "techtimes.com": 10, - "thesun.co.uk": 10, - "boingboing.net": 9, - "foxbusiness.com": 9, - "indiatoday.in": 9, - "interestingengineering.com": 9, - "inverse.com": 9, - "theconversation.com": 9, - "biometricupdate.com": 8, - "cbc.ca": 8, - "extremetech.com": 8, - "fastcompany.com": 8, - "smh.com.au": 8, - "techdirt.com": 8, - "vox.com": 8, - "zdnet.com": 8, - "euronews.com": 7, - "freep.com": 7, - "hindustantimes.com": 7, - "huffingtonpost.com": 7, - "iflscience.com": 7, - "inc.com": 7, - "money.cnn.com": 7, - "newsweek.com": 7, - "nymag.com": 7, - "petapixel.com": 7, - "siliconangle.com": 7, - "splinternews.com": 7, - "stuff.co.nz": 7, - "thedailybeast.com": 7, - "timesnownews.com": 7, - "wionews.com": 7, - "aclu.org": 6, - "algorithmwatch.org": 6, - "aljazeera.com": 6, - "analyticsindiamag.com": 6, - "hrw.org": 6, - "huffpost.com": 6, - "mirror.co.uk": 6, - "news.com.au": 6, - "newser.com": 6, - "newsnationnow.com": 6, - "politico.eu": 6, - "popularmechanics.com": 6, - "statnews.com": 6, - "themarkup.org": 6, - "timesofindia.indiatimes.com": 6, - "wired.co.uk": 6, - "axios.com": 5, - "bbc.co.uk": 5, - "carscoops.com": 5, - "cointelegraph.com": 5, - "cyberscoop.com": 5, - "economist.com": 5, - "en.wikipedia.org": 5, - "entrepreneur.com": 5, - "france24.com": 5, - "ft.com": 5, - "ftc.gov": 5, - "jalopnik.com": 5, - "koreaherald.com": 5, - "metro.co.uk": 5, - "ndtv.com": 5, - "newsguardtech.com": 5, - "papers.ssrn.com": 5, - "readwrite.com": 5, - "scmp.com": 5, - "snopes.com": 5, - "thedrive.com": 5, - "theinquirer.net": 5, - "theintercept.com": 5, - "torquenews.com": 5, - "tribune.com.pk": 5, - "404media.co": 4, - "autoblog.com": 4, - "autoevolution.com": 4, - "autonews.com": 4, - "baijiahao.baidu.com": 4, - "complex.com": 4, - "cybernews.com": 4, - "digitaljournal.com": 4, - "factcheck.afp.com": 4, - "foxglove.org.uk": 4, - "globalnews.ca": 4, - "huffingtonpost.co.uk": 4, - "huffingtonpost.com.au": 4, - "indiatimes.com": 4, - "insideevs.com": 4, - "justice.gov": 4, - "koreatimes.co.kr": 4, - "kotaku.com": 4, - "malwaretips.com": 4, - "mercurynews.com": 4, - "morningbrew.com": 4, - "msn.com": 4, - "nydailynews.com": 4, - "nzherald.co.nz": 4, - "patch.com": 4, - "phys.org": 4, - "privacyinternational.org": 4, - "projects.tampabay.com": 4, - "rt.com": 4, - "sbs.com.au": 4, - "spectrum.ieee.org": 4, - "standard.co.uk": 4, - "static.nhtsa.gov": 4, - "straitstimes.com": 4, - "tapinto.net": 4, - "tech.co": 4, - "techrepublic.com": 4, - "themessenger.com": 4, - "variety.com": 4, - "windowscentral.com": 4, - "youtube.com": 4, - "9to5mac.com": 3, - "about.fb.com": 3, - "al-monitor.com": 3, - "alphr.com": 3, - "billboard.com": 3, - "bizjournals.com": 3, - "blackenterprise.com": 3, - "bostonglobe.com": 3, - "businessinsider.in": 3, - "caranddriver.com": 3, - "chicago.suntimes.com": 3, - "choice.com.au": 3, - "courthousenews.com": 3, - "dailycaller.com": 3, - "dailyo.in": 3, - "democratandchronicle.com": 3, - "dexerto.com": 3, - "dmv.ca.gov": 3, - "eff.org": 3, - "en.yna.co.kr": 3, - "fiercehealthcare.com": 3, - "finance.yahoo.com": 3, - "globalwitness.org": 3, - "haaretz.com": 3, - "hothardware.com": 3, - "ibtimes.co.uk": 3, - "infoworld.com": 3, - "insidehighered.com": 3, - "insider.com": 3, - "insurancejournal.com": 3, - "itechpost.com": 3, - "itnews.com.au": 3, - "jezebel.com": 3, - "lawandcrime.com": 3, - "lifehacker.com": 3, - "linkedin.com": 3, - "livemint.com": 3, - "logicallyfacts.com": 3, - "marketwatch.com": 3, - "mic.com": 3, - "mindmatters.ai": 3, - "nature.com": 3, - "news.bloomberglaw.com": 3, - "news.ycombinator.com": 3, - "news18.com": 3, - "newsbytesapp.com": 3, - "newscientist.com": 3, - "newshub.co.nz": 3, - "nj.com": 3, - "pandaily.com": 3, - "people.com": 3, - "politico.com": 3, - "polygon.com": 3, - "protocol.com": 3, - "reddit.com": 3, - "restofworld.org": 3, - "roboticsbusinessreview.com": 3, - "rollingstone.com": 3, - "sciencealert.com": 3, - "sfexaminer.com": 3, - "sfgate.com": 3, - "siliconrepublic.com": 3, - "sixthtone.com": 3, - "tampabay.com": 3, - "tech.slashdot.org": 3, - "techinasia.com": 3, - "techpolicy.press": 3, - "the-independent.com": 3, - "the-sun.com": 3, - "theautopian.com": 3, - "thedrum.com": 3, - "thehindu.com": 3, - "thestreet.com": 3, - "thewrap.com": 3, - "today.com": 3, - "trustedreviews.com": 3, - "voanews.com": 3, - "washingtontimes.com": 3, - "weforum.org": 3, - "abajournal.com": 2, - "abovethelaw.com": 2, - "acsh.org": 2, - "adcu.org.uk": 2, - "advox.globalvoices.org": 2, - "afr.com": 2, - "ajc.com": 2, - "allure.com": 2, - "amnesty.org": 2, - "androidauthority.com": 2, - "aol.com": 2, - "appleinsider.com": 2, - "article.wn.com": 2, - "au.pcmag.com": 2, - "audacy.com": 2, - "balkaninsight.com": 2, - "blog.google": 2, - "blogs.wsj.com": 2, - "brookings.edu": 2, - "brusselstimes.com": 2, - "businesstoday.in": 2, - "bustle.com": 2, - "canadianlawyermag.com": 2, - "ccn.com": 2, - "cfodive.com": 2, - "chicagomag.com": 2, - "cityandstateny.com": 2, - "coindesk.com": 2, - "computerworld.com.au": 2, - "consumer.org.nz": 2, - "consumerist.com": 2, - "cosmopolitan.com": 2, - "cultofmac.com": 2, - "cyberdaily.au": 2, - "dailynews.com": 2, - "dailywire.com": 2, - "darkreading.com": 2, - "dataconomy.com": 2, - "decrypt.co": 2, - "denver7.com": 2, - "dezeen.com": 2, - "digitalethics.org": 2, - "digitalspy.com": 2, - "disabilitynewsservice.com": 2, - "driveteslacanada.ca": 2, - "driving.co.uk": 2, - "dukechronicle.com": 2, - "edmontonpolice.ca": 2, - "ekker.legal": 2, - "euractiv.com": 2, - "express.co.uk": 2, - "facebook.com": 2, - "financialexpress.com": 2, - "floridapolitics.com": 2, - "fox4news.com": 2, - "frontofficesports.com": 2, - "fullfact.org": 2, - "gearbrain.com": 2, - "geek.com": 2, - "geekwire.com": 2, - "giantfreakinrobot.com": 2, - "github.com": 2, - "glamour.com": 2, - "globalvillagespace.com": 2, - "govtech.com": 2, - "graphika.com": 2, - "hackingdistributed.com": 2, - "hackread.com": 2, - "hai.stanford.edu": 2, - "hani.co.kr": 2, - "hypebeast.com": 2, - "ic3.gov": 2, - "independent.ie": 2, - "inews.co.uk": 2, - "infowars.com": 2, - "inputmag.com": 2, - "inquisitr.com": 2, - "intellinews.com": 2, - "internetofbusiness.com": 2, - "iol.co.za": 2, - "irishtimes.com": 2, - "isdglobal.org": 2, - "jpost.com": 2, - "kffhealthnews.org": 2, - "kiro7.com": 2, - "knightlawgroup.com": 2, - "ktla.com": 2, - "lalibre.be": 2, - "latestly.com": 2, - "lexology.com": 2, - "link.springer.com": 2, - "lsj.com.au": 2, - "mediamatters.org": 2, - "min.news": 2, - "motherjones.com": 2, - "nationalpost.com": 2, - "nationalreview.com": 2, - "nbcbayarea.com": 2, - "news.abplive.com": 2, - "news.trust.org": 2, - "newsone.com": 2, - "newyorker.com": 2, - "nme.com": 2, - "nola.com": 2, - "notebookcheck.net": 2, - "ntsb.gov": 2, - "openai.com": 2, - "pagesix.com": 2, - "pcworld.com": 2, - "pnas.org": 2, - "politifact.com": 2, - "popsci.com": 2, - "post-gazette.com": 2, - "pymnts.com": 2, - "rappler.com": 2, - "reason.com": 2, - "recode.net": 2, - "researchgate.net": 2, - "revealnews.org": 2, - "sacbee.com": 2, - "screenshot-media.com": 2, - "sea.mashable.com": 2, - "seanbmcgregor.com": 2, - "searchengineland.com": 2, - "seattletimes.com": 2, - "sec.gov": 2, - "seekingalpha.com": 2, - "sfchronicle.com": 2, - "sfist.com": 2, - "simonwillison.net": 2, - "sitepronews.com": 2, - "spiegel.de": 2, - "startribune.com": 2, - "taiwannews.com.tw": 2, - "techmonitor.ai": 2, - "techradar.com": 2, - "techspot.com": 2, - "techtransparencyproject.org": 2, - "techxplore.com": 2, - "the-decoder.com": 2, - "thebulletin.org": 2, - "thecut.com": 2, - "thediplomat.com": 2, - "theinformation.com": 2, - "thenewstack.io": 2, - "thequint.com": 2, - "thestar.com": 2, - "tiktok.com": 2, - "tomsguide.com": 2, - "turkishminute.com": 2, - "uk.news.yahoo.com": 2, - "uni-watch.com": 2, - "unilad.com": 2, - "usnews.com": 2, - "vanderbilthustler.com": 2, - "vietnamnet.vn": 2, - "voicebot.ai": 2, - "walesonline.co.uk": 2, - "wral.com": 2, - "youtu.be": 2, - "10news.com": 1, - "11alive.com": 1, - "2025ad.com": 1, - "71republic.com": 1, - "947.co.za": 1, - "972mag.com": 1, - "9news.com": 1, - "9news.com.au": 1, - "9to5google.com": 1, - "aa.com.tr": 1, - "aaai.org": 1, - "aap.com.au": 1, - "abc15.com": 1, - "abc4.com": 1, - "abc7.com": 1, - "abc7news.com": 1, - "abc7ny.com": 1, - "abplive.in": 1, - "accessnow.org": 1, - "aclunc.org": 1, - "advrider.com": 1, - "adweek.com": 1, - "aebsettlement.com": 1, - "afrotech.com": 1, - "agora.md": 1, - "aibusiness.com": 1, - "aisnakeoil.substack.com": 1, - "alaskasnewssource.com": 1, - "allaboutai.com": 1, - "amp-theguardian-com.cdn.ampproject.org": 1, - "analyticsdrift.com": 1, - "analyticsvidhya.com": 1, - "androidheadlines.com": 1, - "antoinespeaks.co.uk": 1, - "aplus.com": 1, - "app.com": 1, - "appleworld.today": 1, - "archdaily.com": 1, - "archive.boston.com": 1, - "artnews.com": 1, - "asiafinancial.com": 1, - "atlanticcouncil.org": 1, - "au.finance.yahoo.com": 1, - "augustman.com": 1, - "autoguide.com": 1, - "autonews.gasgoo.com": 1, - "autosafety.org": 1, - "awfulannouncing.com": 1, - "azcentral.com": 1, - "azfamily.com": 1, - "bair.berkeley.edu": 1, - "bangkokpost.com": 1, - "bankinfosecurity.com": 1, - "barrysookman.com": 1, - "barstoolsports.com": 1, - "bartlesvilleradio.com": 1, - "beckersasc.com": 1, - "beckershospitalreview.com": 1, - "beincrypto.com": 1, - "belairassociation.org": 1, - "belganewsagency.eu": 1, - "benitolink.com": 1, - "benzinga.com": 1, - "biggerlawfirm.com": 1, - "bipartisanpolicy.org": 1, - "bitdefender.com": 1, - "bits.blogs.nytimes.com": 1, - "blavity.com": 1, - "blitzquotidiano.it": 1, - "blockchain-council.org": 1, - "blockchaintechnology-news.com": 1, - "blog.angryasianman.com": 1, - "blog.bity.com": 1, - "blog.botego.com": 1, - "blog.coinbase.com": 1, - "blog.conceptnet.io": 1, - "blog.galalaw.com": 1, - "blog.openai.com": 1, - "blog.slock.it": 1, - "blog.twitter.com": 1, - "blogs.bing.com": 1, - "blogs.discovermagazine.com": 1, - "blogs.microsoft.com": 1, - "bloomberg.com.": 1, - "bnnbloomberg.ca": 1, - "bnnbreaking.com": 1, - "boredpanda.com": 1, - "boston.com": 1, - "bostonherald.com": 1, - "boxmining.com": 1, - "br.de": 1, - "brecorder.com": 1, - "breitbart.com": 1, - "bronx.news12.com": 1, - "business-standard.com": 1, - "businesscloud.co.uk": 1, - "businessnewsdaily.com": 1, - "businesstimes.com.sg": 1, - "businesswire.com": 1, - "bussgeldportal.de": 1, - "buzzfeed.com": 1, - "ca.finance.yahoo.com": 1, - "ca.movies.yahoo.com": 1, - "ca.news.yahoo.com": 1, - "cacm.acm.org": 1, - "cafemom.com": 1, - "caixinglobal.com": 1, - "capitalbrief.com": 1, - "capradio.org": 1, - "carcomplaints.com": 1, - "carloscreusmoreira.medium.com": 1, - "carthrottle.com": 1, - "casetext.com": 1, - "catonetworks.com": 1, - "cbs12.com": 1, - "cbs17.com": 1, - "ccjdigital.com": 1, - "cdn.arstechnica.net": 1, - "cdn.openai.com": 1, - "cdn.sanity.io": 1, - "cdotrends.com": 1, - "cdt.org": 1, - "cepa.org": 1, - "cfr.org": 1, - "cgdev.org": 1, - "change.org": 1, - "channel4.com": 1, - "channels.theinnovationenterprise.com": 1, - "chatbotslife.com": 1, - "chatbotsmagazine.com": 1, - "chicagoreader.com": 1, - "chicagotribune.com": 1, - "chinadaily.com.cn": 1, - "christiantoday.com": 1, - "cinemablend.com": 1, - "cio.com": 1, - "cityam.com": 1, - "citylab.com": 1, - "civilbeat.org": 1, - "cjr.org": 1, - "clarin.com": 1, - "clarksonlawfirm.com": 1, - "claytonutz.com": 1, - "click2houston.com": 1, - "cmu.edu": 1, - "codastory.com": 1, - "cohenmilstein.com": 1, - "coincentral.com": 1, - "coincodex.com": 1, - "coinmarketcap.com": 1, - "colombiaone.com": 1, - "coloradohometownweekly.com": 1, - "colorlines.com": 1, - "commondreams.org": 1, - "comptroller.nyc.gov": 1, - "computerworld.com": 1, - "computing.co.uk": 1, - "consumersinternational.org": 1, - "context-cdn.washingtonpost.com": 1, - "copyrightlately.com": 1, - "corporatefinanceinstitute.com": 1, - "cosmeticsbusiness.com": 1, - "cosmosmagazine.com": 1, - "coyotecountrylv.com": 1, - "cpomagazine.com": 1, - "cpsc.gov": 1, - "cracked.com": 1, - "creativebloq.com": 1, - "cryptocompare.com": 1, - "cryptocurrencyhub.io": 1, - "cryptopolitan.com": 1, - "csmapnyu.org": 1, - "csoonline.com": 1, - "ctinsider.com": 1, - "ctvnews.ca": 1, - "culvercityobserver.com": 1, - "curiosity.com": 1, - "cyber.fsi.stanford.edu": 1, - "cybersecurityfordemocracy.org": 1, - "d.dailynews.co.th": 1, - "dailybreeze.com": 1, - "dailycitizen.focusonthefamily.com": 1, - "dailypress.net": 1, - "dailystar.co.uk": 1, - "dallasnews.com": 1, - "danrl.com": 1, - "dataprivacylab.org": 1, - "deadline.com": 1, - "dealstreetasia.com": 1, - "dearauthor.com": 1, - "deccanherald.com": 1, - "delish.com": 1, - "denverpost.com": 1, - "designnews.com": 1, - "detroitnews.com": 1, - "digg.com": 1, - "digit.fyi": 1, - "digitalcameraworld.com": 1, - "dispatch.com": 1, - "disputeresolutiongermany.com": 1, - "dkb.blog": 1, - "dl.acm.org": 1, - "dmlp.org": 1, - "dmv.com": 1, - "dnaindia.com": 1, - "docs.rwu.edu": 1, - "donga.com": 1, - "dotesports.com": 1, - "dpreview.com": 1, - "dttc.sggp.org.vn": 1, - "dw.com": 1, - "ebaumsworld.com": 1, - "ecns.cn": 1, - "economictimes.indiatimes.com": 1, - "econotimes.com": 1, - "editorialge.com": 1, - "edpb.europa.eu": 1, - "edrants.com": 1, - "eldiario.es": 1, - "electronicproducts.com": 1, - "elitereaders.com": 1, - "elle.com": 1, - "elmundo.es": 1, - "elobservador.com.uy": 1, - "emirates247.com": 1, - "en.dailypakistan.com.pk": 1, - "en.econostrum.info": 1, - "en.wikinews.org": 1, - "energyinfrapost.com": 1, - "english.elpais.com": 1, - "english.jagran.com": 1, - "ent.siteintelgroup.com": 1, - "epic.org": 1, - "escapistmagazine.com": 1, - "eteknix.com": 1, - "etftrends.com": 1, - "eticasfoundation.org": 1, - "eu.usatoday.com": 1, - "eulawenforcement.com": 1, - "eurogamer.net": 1, - "europepmc.org": 1, - "explica.co": 1, - "expressnews.com": 1, - "fairfieldsuntimes.com": 1, - "fairworkweek.org": 1, - "familyhandyman.com": 1, - "features.propublica.org": 1, - "finance.nine.com.au": 1, - "financialregnews.com": 1, - "finbold.com": 1, - "firstpost.com": 1, - "flawedfacedata.com": 1, - "fool.com": 1, - "forbesafrica.com": 1, - "fordfoundation.org": 1, - "foreignpolicy.com": 1, - "forum.facepunch.com": 1, - "forward.com": 1, - "fossa.com": 1, - "foundation.mozilla.org": 1, - "fox2detroit.com": 1, - "fox5dc.com": 1, - "fox5ny.com": 1, - "fox5sandiego.com": 1, - "foxcarolina.com": 1, - "freakonomics.com": 1, - "freemovement.org.uk": 1, - "fudzilla.com": 1, - "fullerproject.org": 1, - "futureoflife.org": 1, - "futureparty.com": 1, - "gamasutra.com": 1, - "gamedaily.biz": 1, - "gamespot.com": 1, - "gameworldobserver.com": 1, - "garyrubinstein.teachforus.org": 1, - "gawker.com": 1, - "geekologie.com": 1, - "gffbrokers.com": 1, - "gineersnow.com": 1, - "github.blog": 1, - "githubcopilotlitigation.com": 1, - "gizchina.com": 1, - "gizmodo.co.uk": 1, - "glamourmagazine.co.uk": 1, - "globalcitizen.org": 1, - "globalnation.inquirer.net": 1, - "globaltimes.cn": 1, - "globalvoices.org": 1, - "gmauthority.com": 1, - "gmw.cn": 1, - "gofundme.com": 1, - "goodmorningamerica.com": 1, - "gothamist.com": 1, - "gov.uk": 1, - "gq.com": 1, - "graphics.wsj.com": 1, - "graziadaily.co.uk": 1, - "greekreporter.com": 1, - "greenbot.com": 1, - "gwern.net": 1, - "gzeromedia.com": 1, - "hackaday.com": 1, - "hackernoon.com": 1, - "hardworkingtrucks.com": 1, - "hawaiinewsnow.com": 1, - "hawaiitribune-herald.com": 1, - "hbr.org": 1, - "hbswk.hbs.edu": 1, - "hcamag.com": 1, - "healthcarefinancenews.com": 1, - "heise.de": 1, - "helpnetsecurity.com": 1, - "heraldnews.com": 1, - "hindi.news18.com": 1, - "hoit.uk": 1, - "home.bt.com": 1, - "hotelmanagement.net": 1, - "howtogeek.com": 1, - "hrmonline.com.au": 1, - "hrtechnologynews.com": 1, - "hstoday.us": 1, - "huffingtonpost.ca": 1, - "huntonprivacyblog.com": 1, - "hurriyetdailynews.com": 1, - "hypebae.com": 1, - "ia.acs.org.au": 1, - "iafrica.com": 1, - "iafrikan.com": 1, - "ibtimes.com": 1, - "idiallo.com": 1, - "ilfattoquotidiano.it": 1, - "ilpost.it": 1, - "imd.org": 1, - "imore.com": 1, - "in.finance.yahoo.com": 1, - "in.mashable.com": 1, - "incidentdatabase.ai": 1, - "india.com": 1, - "indianexpress.com": 1, - "indiatvnews.com": 1, - "industryweek.com": 1, - "indystar.com": 1, - "information-age.com": 1, - "innotechtoday.com": 1, - "inshorts.com": 1, - "insideedition.com": 1, - "insidehook.com": 1, - "insideprivacy.com": 1, - "insights.globalspec.com": 1, - "insights.tmpw.co.uk": 1, - "instagram.com": 1, - "intelligenttransport.com": 1, - "intouchweekly.com": 1, - "investors.com": 1, - "io9.gizmodo.com": 1, - "iottechnews.com": 1, - "iotworldtoday.com": 1, - "iphonehacks.com": 1, - "ipi.media": 1, - "ipn.md": 1, - "ipvm.com": 1, - "irishmirror.ie": 1, - "ishn.com": 1, - "itwire.com": 1, - "ivebeenmugged.typepad.com": 1, - "jcwi.org.uk": 1, - "jdsupra.com": 1, - "jsjc.gov.cn": 1, - "jta.org": 1, - "judiciary.uk": 1, - "justcareusa.org": 1, - "justjared.com": 1, - "justsomething.co": 1, - "k.sina.cn": 1, - "kathmandupost.ekantipur.com": 1, - "kfm.co.za": 1, - "khaleejtimes.com": 1, - "khq.com": 1, - "kingsleynapley.co.uk": 1, - "knowyourmeme.com": 1, - "kotaku.co.uk": 1, - "kr-asia.com": 1, - "krdo.com": 1, - "krem.com": 1, - "kslnewsradio.com": 1, - "ktvu.com": 1, - "kurdistan24.net": 1, - "labsnews.com": 1, - "ladbible.com": 1, - "lagrandeobserver.com": 1, - "laion.ai": 1, - "lamag.com": 1, - "lasvegassun.com": 1, - "latimesblogs.latimes.com": 1, - "latintimes.com": 1, - "laweekly.com": 1, - "lawenforcementtoday.com": 1, - "lawfaremedia.org": 1, - "lawgazette.co.uk": 1, - "learningenglish.voanews.com": 1, - "legalreader.com": 1, - "lemonde.fr": 1, - "lentrepreneur.co": 1, - "lesswrong.com": 1, - "lgbtqnation.com": 1, - "liaa.dc.uba.ar": 1, - "lifesitenews.com": 1, - "livescience.com": 1, - "lobste.rs": 1, - "local12.com": 1, - "losangelesblade.com": 1, - "lr21.com.uy": 1, - "macarthurjustice.org": 1, - "macrumors.com": 1, - "madisonrecord.com": 1, - "makeuseof.com": 1, - "malaymail.com": 1, - "malicious.life": 1, - "mansworldindia.com": 1, - "marginalrevolution.com": 1, - "marketplace.org": 1, - "markey.senate.gov": 1, - "martinmontilino.com": 1, - "massdevice.com": 1, - "massivelyop.com": 1, - "mcknightshomecare.com": 1, - "mcvuk.com": 1, - "me.pcmag.com": 1, - "mediaite.com": 1, - "medianama.com": 1, - "medicaleconomics.com": 1, - "medium.freecodecamp.org": 1, - "medscape.com": 1, - "mentalfloss.com": 1, - "meta-writer.livejournal.com": 1, - "meta.stackoverflow.com": 1, - "metalinjection.net": 1, - "metrorailnews.in": 1, - "metroweekly.com": 1, - "meyerweb.com": 1, - "miamiherald.com": 1, - "michaeleisen.org": 1, - "michiganradio.org": 1, - "middleeasteye.net": 1, - "military.com": 1, - "mlive.com": 1, - "mobility21.cmu.edu": 1, - "mondaq.com": 1, - "montevideo.com.uy": 1, - "montrealgazette.com": 1, - "mothership.sg": 1, - "motorbiscuit.com": 1, - "motormoutharabia.com": 1, - "motortrend.com": 1, - "msnbc.com": 1, - "multichain.com": 1, - "musically.com": 1, - "muycomputerpro.com": 1, - "myfox28columbus.com": 1, - "mymodernmet.com": 1, - "mynorthwest.com": 1, - "nafcu.org": 1, - "nasdaq.com": 1, - "natbuckley.co.uk": 1, - "nation.com.pk": 1, - "natlawreview.com": 1, - "nautil.us": 1, - "nbc15.com": 1, - "nbcboston.com": 1, - "nbcchicago.com": 1, - "nbcdfw.com": 1, - "nbclosangeles.com": 1, - "nbcmiami.com": 1, - "nbcnewyork.com": 1, - "nbcphiladelphia.com": 1, - "nbcwashington.com": 1, - "nec.com": 1, - "neil-clarke.com": 1, - "neoskosmos.com": 1, - "neowin.net": 1, - "newindianexpress.com": 1, - "newjerseymonitor.com": 1, - "newmatilda.com": 1, - "newrepublic.com": 1, - "news.avclub.com": 1, - "news.cgtn.com": 1, - "news.cornell.edu": 1, - "news.trendmicro.com": 1, - "news.tvbs.com.tw": 1, - "news.yahoo.com": 1, - "news5cleveland.com": 1, - "news9live.com": 1, - "newsbeezer.com": 1, - "newsbtc.com": 1, - "newsfolo.com": 1, - "newsmax.com": 1, - "newsobserver.com": 1, - "newson6.com": 1, - "newsroom.gettyimages.com": 1, - "newstatesman.com": 1, - "newswise.com": 1, - "newsx.com": 1, - "nguoi-viet.com": 1, - "nickdiakopoulos.com": 1, - "nj1015.com": 1, - "nknews.org": 1, - "northjersey.com": 1, - "nouse.co.uk": 1, - "nowthisnews.com": 1, - "nst.com.my": 1, - "ntd.com": 1, - "ny1.com": 1, - "observers.france24.com": 1, - "ocregister.com": 1, - "offthemainpage.com": 1, - "olhardigital.com.br": 1, - "opb.org": 1, - "osf.io": 1, - "out-law.com": 1, - "overlayfalseclaims.com": 1, - "ovic.vic.gov.au": 1, - "palestinechronicle.com": 1, - "pantip.com": 1, - "parentadvocates.org": 1, - "pastemagazine.com": 1, - "pbs.org": 1, - "pcgamer.com": 1, - "pewtrusts.org": 1, - "phonearena.com": 1, - "physicstoday.scitation.org": 1, - "plagiarismtoday.com": 1, - "plantservices.com": 1, - "pocket-lint.com": 1, - "policyoptions.irpp.org": 1, - "poliisi.fi": 1, - "polizei.bayern.de": 1, - "polyesterstudio.com": 1, - "populardemocracy.org": 1, - "postcourier.com.pg": 1, - "poynter.org": 1, - "pressgazette.co.uk": 1, - "pride.com": 1, - "princeton.edu": 1, - "priv.gc.ca": 1, - "prnewswire.com": 1, - "proactiveinvestors.com": 1, - "probonoaustralia.com.au": 1, - "projects.propublica.org": 1, - "psmag.com": 1, - "psychiatrist.com": 1, - "publiclawproject.org.uk": 1, - "pulitzercenter.org": 1, - "pulsenews.co.kr": 1, - "punchbowl.news": 1, - "punto-informatico.it": 1, - "qcnews.com": 1, - "queerinai.com": 1, - "queerty.com": 1, - "queue.acm.org": 1, - "quinyx.com": 1, - "radionz.co.nz": 1, - "raillc.substack.com": 1, - "rand.org": 1, - "read.dukeupress.edu": 1, - "readlion.com": 1, - "recordedfuture.com": 1, - "republicworld.com": 1, - "research.checkpoint.com": 1, - "resemble.ai": 1, - "restaurantbusinessonline.com": 1, - "retailwire.com": 1, - "retractionwatch.com": 1, - "rsk.co": 1, - "rss.onlinelibrary.wiley.com": 1, - "rts.ch": 1, - "rtve.es": 1, - "russiamatters.org": 1, - "s.weibo.com": 1, - "safiyaunoble.files.wordpress.com": 1, - "salon.com": 1, - "samaa.tv": 1, - "sbnation.com": 1, - "schneier.com": 1, - "science.org": 1 - }, - "CSETv1.AI Task": { - "facial recognition": 131, - "navigation": 112, - "patrolling": 62, - "virtual assistant technology": 58, - "security": 57, - "object detection": 55, - "autonomous driving": 54, - "self-driving": 53, - "object recognition": 51, - "chat bot": 49, - "text generation": 46, - "search optimization": 44, - "autonomous navigation": 43, - "search suggestion": 41, - "Image search": 40, - "predict recidivism": 38, - "Rank Applicants": 37, - "translation": 37, - "prediction": 35, - "self driving": 35, - "semi-autonomous navigation": 35, - "surveillance": 35, - "Driving": 34, - "resume screening": 33, - "content generation": 31, - "deepfake video generation": 31, - "autocomplete": 30, - "content moderation": 30, - "recommender": 30, - "content ranking": 29, - "Generate Captions": 28, - "assembly": 28, - "image identification": 28, - "image interpretation": 28, - "production": 28, - "personalized online advertising": 27, - "image classification": 25, - "jaywalking detection": 25, - "Autopilot": 24, - "Image Tagging": 24, - "identity verification": 24, - "image categorization": 24, - "obstacle avoidance": 24, - "route optimization": 24, - "filter": 23, - "image modification": 23, - "photo edit": 23, - "Generate Replies": 22, - "chatbot": 22, - "face detection": 22, - "search engine optimization": 19, - "Predict Crimes": 13, - "predictive policing": 13, - "recommendation": 13, - "generation": 12, - "detect gunshots": 11, - "identify weapon calibre": 11, - "locate gunshots": 11, - "personalized online search results": 11, - "population of characteristics for NPCs in a video game": 11, - "predict shooter movement": 11, - "Image Analysis": 10, - "predict store traffic": 10, - "productivity optimization": 10, - "scheduling": 10, - "image generation": 9, - "toxicity detection": 9, - "voice recognition": 9, - "augmented reality (AR) game Pokémon Go": 8, - "natural language processing": 8, - "speech recognition": 8, - "Design Phone Cases": 7, - "assign risk": 7, - "classification": 7, - "deepfake audio generation": 7, - "predict healthcare needs": 7, - "sentiment analysis": 7, - "editing": 6, - "fraud risk prediction": 6, - "identify hate speech": 6, - "webpage maintenance": 6, - "Crop Images": 5, - "Personal Assistant": 5, - "automatic emergency braking": 5, - "content recommendation": 5, - "image cropping": 5, - "security monitor": 5, - "speech interpretation": 5, - "Digital Assistant": 4, - "application screening": 4, - "audit selection": 4, - "candidate assessment": 4, - "diagnose sepsis": 4, - "emotion recognition": 4, - "estimating house prices": 4, - "facial expression recognition": 4, - "home valuation": 4, - "identify sepsis": 4, - "object classification": 4, - "price determination": 4, - "product promotion": 4, - "real estate market forecasting": 4, - "search engine": 4, - "search result ranking": 4, - "sepsis prediction": 4, - "transportation": 4, - "virtual interview": 4, - "worker management": 4, - "delivery": 3, - "ethical decision making": 3, - "gender classification": 3, - "move shelves": 3, - "performance tracking": 3, - "reply to questions": 3, - "video recommender": 3, - "water quality testing": 3, - "worker monitoring": 3, - "assess applicants": 2, - "audio deepfake": 2, - "automated speech recognition": 2, - "automatic tracking": 2, - "ball detection": 2, - "chant scripture": 2, - "cheerleading": 2, - "concierge": 2, - "driver surveillance": 2, - "emotion detection": 2, - "entertain guests": 2, - "event detection": 2, - "event prediction": 2, - "exercise demonstration": 2, - "eye tracking": 2, - "facial grouping": 2, - "household administration": 2, - "identification or detection": 2, - "product recognition": 2, - "rank the reliability of workers": 2, - "risk assessment": 2, - "safe driving detection": 2, - "semi-autonomous driving": 2, - "shift assignment": 2, - "speech-to-text": 2, - "suggestion": 2, - "weapon detection": 2, - "welfare determination": 2, - "ADA website compliance": 1, - "Burmese to English translation": 1, - "Matching": 1, - "accessibility compliance": 1, - "advertisement screening": 1, - "autonomous food delivery": 1, - "benefits allocation": 1, - "brand safety detection": 1, - "content suggestion": 1, - "customer service": 1, - "database": 1, - "detect policy violations": 1, - "enforcement of community guidelines": 1, - "engagement optimization": 1, - "exam proctoring": 1, - "fact-check": 1, - "human facial image transformation": 1, - "identify misinformation": 1, - "image editing": 1, - "image organization": 1, - "image recognition": 1, - "image splicing": 1, - "job screening": 1, - "law enforcement": 1, - "license plate matching": 1, - "license plate recognition": 1, - "marketing material generation": 1, - "optimize movement to win a computer game": 1, - "passport photo quality check": 1, - "performance prediction": 1, - "predict grades on exam": 1, - "predict necessary amount of medical coverage": 1, - "predict recidivism risk": 1, - "recipe generation": 1, - "recruitment screening": 1, - "remote proctoring": 1, - "risk prediction": 1, - "robotics": 1, - "spam filter": 1, - "stoplight recognition": 1, - "vaccine allocation": 1, - "video classification": 1, - "video suggestion": 1, - "voice alteration": 1, - "weapons detection": 1, - "word association": 1 - }, - "CSETv1.Deployed": { - "yes": 1442, - "no": 103, - "maybe": 37 - }, - "CSETv1.Embedded": { - "no": 976, - "yes": 592, - "maybe": 14 - }, - "CSETv1.Injuries": { - "0": 1480, - "1": 66, - "2": 5, - "10": 27, - "40": 1, - "54": 17, - "416": 3, - "1391": 12, - "55000": 3 - }, - "classifications": { - "CSETv1_Annotator-1:User Test in Controlled Conditions:no": 1841, - "CSETv1_Annotator-1:Impact on Critical Services:no": 1799, - "CSETv1_Annotator-1:Multiple AI Interaction:no": 1797, - "CSETv1_Annotator-1:Producer Test in Controlled Conditions:no": 1785, - "CSETv1_Annotator-1:Report, Test, or Study of data:no": 1755, - "CSETv1_Annotator-1:Deployed:yes": 1715, - "CSETv1_Annotator-1:Involving Minor:no": 1701, - "CSETv1_Annotator-1:Rights Violation:no": 1690, - "CSETv1_Annotator-1:Producer Test in Operational Conditions:no": 1683, - "CSETv1_Annotator-1:User Test in Operational Conditions:no": 1567, - "CSETv1:User Test in Controlled Conditions:no": 1555, - "CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm": 1542, - "CSETv1_Annotator-1:Public Sector Deployment:no": 1526, - "CSETv1:Multiple AI Interaction:no": 1525, - "CSETv1:Report, Test, or Study of data:no": 1511, - "CSETv1:Producer Test in Controlled Conditions:no": 1509, - "CSETv1_Annotator-1:Detrimental Content:no": 1499, - "CSETv1_Annotator-1:Entertainment Industry:no": 1498, - "CSETv1_Annotator-1:AI System:yes": 1497, - "CSETv1:Impact on Critical Services:no": 1474, - "CSETv1:Involving Minor:no": 1469, - "CSETv1:Intentional Harm:No. Not intentionally designed to perform harm": 1462, - "CSETv1:Producer Test in Operational Conditions:no": 1453, - "CSETv1:Deployed:yes": 1442, - "CSETv1:Clear link to Technology:yes": 1425, - "CSETv1_Annotator-1:Clear link to Technology:yes": 1422, - "CSETv1:Entertainment Industry:no": 1401, - "CSETv1:User Test in Operational Conditions:no": 1367, - "CSETv1:Rights Violation:no": 1332, - "CSETv1:Public Sector Deployment:no": 1324, - "CSETv1:Detrimental Content:no": 1281, - "CSETv1_Annotator-1:Protected Characteristic:no": 1270, - "CSETv1_Annotator-1:Harm Distribution Basis:none": 1223, - "CSETv1:AI System:yes": 1221, - "CSETv1_Annotator-1:Embedded:no": 1206, - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no": 1192, - "CSETv1_Annotator-1:AI Harm Level:none": 1153, - "CSETv1_Annotator-1:Physical Objects:no": 1150, - "CSETv1_Annotator-1:Special Interest Intangible Harm:no": 1087, - "CSETv1:Protected Characteristic:no": 1070, - "CSETv1:Harm Distribution Basis:none": 1059, - "CSETv1:AI Harm Level:none": 1003, - "CSETv1:Embedded:no": 976, - "CSETv1:Annotator’s AI special interest intangible harm assessment:no": 949, - "CSETv1:Physical Objects:no": 946, - "CSETv1_Annotator-1:Location Country (two letters):US": 946, - "CSETv0:Intent:Accident": 935, - "CSETv1_Annotator-2:User Test in Controlled Conditions:no": 930, - "CSETv0:Relevant AI functions:Cognition": 924, - "CSETv1_Annotator-2:Entertainment Industry:no": 908, - "CSETv1_Annotator-2:Multiple AI Interaction:no": 904, - "CSETv1_Annotator-2:Producer Test in Operational Conditions:no": 904, - "CSETv1_Annotator-2:Impact on Critical Services:no": 901, - "CSETv1_Annotator-1:Location Region:North America": 899, - "CSETv0:Nature of End User:Amateur": 896, - "CSETv0:Relevant AI functions:Perception": 892, - "CSETv1_Annotator-2:Producer Test in Controlled Conditions:no": 884, - "CSETv1_Annotator-2:Report, Test, or Study of data:no": 870, - "CSETv1:Special Interest Intangible Harm:no": 862, - "CSETv1_Annotator-1:Autonomy Level:Autonomy1": 852, - "CSETv1_Annotator-2:Intentional Harm:No. Not intentionally designed to perform harm": 850, - "CSETv1:Autonomy Level:Autonomy1": 847, - "CSETv1_Annotator-2:Involving Minor:no": 843, - "CSETv1_Annotator-2:Deployed:yes": 841, - "CSETv1_Annotator-2:Rights Violation:no": 826, - "CSETv1_Annotator-2:Public Sector Deployment:no": 819, - "CSETv1_Annotator-2:Detrimental Content:no": 813, - "CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue": 794, - "CSETv1:Location Region:North America": 781, - "CSETv1_Annotator-2:User Test in Operational Conditions:no": 773, - "CSETv1:Location Country (two letters):US": 769, - "CSETv0:Relevant AI functions:Action": 768, - "CSETv1_Annotator-1:Special Interest Intangible Harm:yes": 732, - "CSETv1_Annotator-1:Tangible Harm:tangible harm definitively occurred": 719, - "CSETv0:Physical System:Software only": 704, - "CSETv1_Annotator-2:AI Harm Level:none": 704, - "CSETv1:Special Interest Intangible Harm:yes": 703, - "CSETv1_Annotator-2:Protected Characteristic:no": 689, - "CSETv1:Tangible Harm:tangible harm definitively occurred": 678, - "CSETv1_Annotator-2:AI System:yes": 675, - "CSETv1_Annotator-1:Physical Objects:yes": 664, - "CSETv1_Annotator-2:Embedded:no": 649, - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:yes": 647, - "CSETv1_Annotator-1:Embedded:yes": 643, - "CSETv1_Annotator-1:Sector of Deployment:information and communication": 639, - "CSETv0:Severity:Negligible": 634, - "CSETv1:Tangible Harm:no tangible harm, near-miss, or issue": 618, - "CSETv1:Physical Objects:yes": 615, - "CSETv1:Annotator’s AI special interest intangible harm assessment:yes": 609, - "CSETv1_Annotator-2:Annotator’s AI special interest intangible harm assessment:no": 607, - "CSETv1_Annotator-2:Harm Distribution Basis:none": 596, - "CSETv1:Embedded:yes": 592, - "CSETv1_Annotator-1:Protected Characteristic:yes": 565, - "CSETv1:Sector of Deployment:information and communication": 562, - "CSETv1_Annotator-2:Special Interest Intangible Harm:no": 551, - "CSETv1_Annotator-2:Physical Objects:no": 550, - "CSETv1_Annotator-1:Autonomy Level:Autonomy3": 546, - "CSETv0:Near Miss:Harm caused": 537, - "CSETv1_Annotator-2:Clear link to Technology:yes": 522, - "CSETv1_Annotator-1:Location Region:Global": 519, - "CSETv0:Problem Nature:Robustness": 507, - "CSETv1:Protected Characteristic:yes": 505, - "CSETv0:Problem Nature:Specification": 473, - "CSETv0:Near Miss:Unclear/unknown": 461, - "CSETv1_Annotator-2:Location Region:North America": 454, - "CSETv0:Level of Autonomy:High": 452, - "CSETv1:Data Inputs:text": 443, - "CSETv1_Annotator-2:Location Country (two letters):US": 441, - "CSETv1_Annotator-3:Producer Test in Controlled Conditions:no": 435, - "CSETv1_Annotator-3:Involving Minor:no": 427, - "CSETv1_Annotator-3:User Test in Controlled Conditions:no": 424, - "CSETv1_Annotator-2:Tangible Harm:no tangible harm, near-miss, or issue": 422, - "CSETv1_Annotator-3:Producer Test in Operational Conditions:no": 420, - "CSETv1_Annotator-3:Multiple AI Interaction:no": 418, - "CSETv1_Annotator-2:Clear link to Technology:no": 414, - "CSETv0:Level of Autonomy:Medium": 413, - "CSETv1_Annotator-3:Report, Test, or Study of data:no": 412, - "CSETv1_Annotator-3:Deployed:yes": 409, - "CSETv1_Annotator-2:Physical Objects:yes": 404, - "CSETv1:Location Region:Global": 402, - "CSETv1_Annotator-3:Detrimental Content:no": 394, - "CSETv1_Annotator-3:Entertainment Industry:no": 391, - "CSETv1_Annotator-1:Date of Incident Year:2017": 389, - "CSETv1_Annotator-2:Special Interest Intangible Harm:yes": 389, - "CSETv1_Annotator-3:Impact on Critical Services:no": 388, - "CSETv1_Annotator-3:Intentional Harm:No. Not intentionally designed to perform harm": 388, - "CSETv1_Annotator-2:Tangible Harm:tangible harm definitively occurred": 383, - "CSETv1:Date of Incident Year:2017": 382, - "CSETv1_Annotator-1:AI Harm Level:AI tangible harm event": 381, - "CSETv1_Annotator-2:Autonomy Level:Autonomy1": 379, - "CSETv1:AI Harm Level:AI tangible harm event": 373, - "CSETv1:Autonomy Level:Autonomy3": 370, - "CSETv1_Annotator-1:Autonomy Level:Autonomy2": 369, - "CSETv1_Annotator-1:Sector of Deployment:transportation and storage": 368, - "CSETv1_Annotator-1:Harm Distribution Basis:race": 364, - "CSETv0:Location:Global": 355, - "CSETv0:Sector of Deployment:Information and communication": 352, - "CSETv1_Annotator-3:AI System:yes": 350, - "CSETv1:Harm Distribution Basis:race": 346, - "CSETv1_Annotator-3:Public Sector Deployment:no": 346, - "CSETv1_Annotator-3:User Test in Operational Conditions:no": 342, - "CSETv1_Annotator-1:Clear link to Technology:no": 339, - "CSETv1_Annotator-2:Annotator’s AI special interest intangible harm assessment:yes": 333, - "CSETv1_Annotator-3:Rights Violation:no": 322, - "CSETv1_Annotator-3:Clear link to Technology:yes": 305, - "CSETv1_Annotator-1:Public Sector Deployment:yes": 304, - "CSETv1_Annotator-2:Embedded:yes": 301, - "CSETv0:Problem Nature:Unknown/unclear": 292, - "CSETv1:Data Inputs:images": 286, - "CSETv1_Annotator-1:AI System:no": 286, - "CSETv1_Annotator-1:Data Inputs:images": 283, - "CSETv1_Annotator-1:User Test in Operational Conditions:yes": 283, - "CSETv0:Harm Type:Harm to physical health/safety": 282, - "CSETv1_Annotator-2:Autonomy Level:Autonomy2": 281, - "CSETv1_Annotator-1:Detrimental Content:yes": 279, - "CSETv1:Sector of Deployment:transportation and storage": 273, - "CSETv1:AI System:no": 269, - "CSETv1_Annotator-2:Sector of Deployment:information and communication": 269, - "CSETv1:Date of Incident Year:2016": 268, - "CSETv1_Annotator-2:Date of Incident Year:2017": 267, - "CSETv0:Severity:Minor": 262, - "CSETv1:Detrimental Content:yes": 262, - "CSETv1:Autonomy Level:Autonomy2": 261, - "CSETv1_Annotator-2:Protected Characteristic:yes": 261, - "CSETv1_Annotator-2:Data Inputs:Text": 260, - "CSETv1_Annotator-3:Protected Characteristic:no": 258, - "CSETv1_Annotator-3:Physical Objects:no": 256, - "CSETv1_Annotator-1:Sector of Deployment:Arts, entertainment and recreation": 252, - "CSETv1_Annotator-3:Harm Distribution Basis:none": 252, - "CSETv1:Public Sector Deployment:yes": 249, - "CSETv1_Annotator-1:Harm Distribution Basis:sex": 243, - "CSETv0:Named Entities:Google": 242, - "CSETv1_Annotator-3:Special Interest Intangible Harm:no": 241, - "CSETv0:Harm Type:Harm to social or political systems": 239, - "CSETv1_Annotator-1:Data Inputs:sensor data": 239, - "CSETv0:Technology Purveyor:Google": 238, - "CSETv0:Harm Distribution Basis:Race": 235, - "CSETv1:AI tools and methods:natural language processing": 235, - "CSETv1:Data Inputs:video": 234, - "CSETv1_Annotator-1:Entertainment Industry:yes": 229, - "CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:no": 226, - "CSETv0:System Developer:Google": 224, - "CSETv1_Annotator-1:Data Inputs:text": 223, - "CSETv1_Annotator-1:Date of Incident Year:2016": 223, - "CSETv1_Annotator-1:Date of Incident Year:2018": 222, - "CSETv0:AI Techniques:Machine learning": 221, - "CSETv1_Annotator-2:Autonomy Level:Autonomy3": 221, - "CSETv0:Harm Type:Financial harm": 220, - "CSETv0:Near Miss:Near miss": 219, - "CSETv1_Annotator-3:Autonomy Level:Autonomy1": 219, - "CSETv1_Annotator-1:Sector of Deployment:law enforcement": 218, - "CSETv1:AI tools and methods:computer vision": 217, - "CSETv1_Annotator-1:Data Inputs:video input": 216, - "CSETv0:Physical System:Vehicle/mobile robot": 215, - "CSETv1_Annotator-2:AI System:no": 215, - "CSETv1_Annotator-3:Embedded:yes": 213, - "CSETv1:User Test in Operational Conditions:yes": 211, - "CSETv1_Annotator-3:Embedded:no": 211, - "CSETv0:Problem Nature:Assurance": 207, - "CSETv1_Annotator-2:Location Region:Global": 205, - "CSETv1_Annotator-3:AI Harm Level:none": 204, - "CSETv0:Harm Type:Psychological harm": 202, - "CSETv1_Annotator-1:Tangible Harm:imminent risk of tangible harm (near miss) did occur": 201, - "CSETv1:Harm Distribution Basis:sex": 198, - "CSETv0:Intent:Unclear": 197, - "CSETv1_Annotator-2:AI tools and methods:Natural Language Processesing": 196, - "CSETv0:Sector of Deployment:Transportation and storage": 191, - "CSETv1:Sector of Deployment:Arts, entertainment and recreation": 190, - "CSETv1_Annotator-3:Location Region:North America": 187, - "CSETv1_Annotator-3:Tangible Harm:tangible harm definitively occurred": 187, - "CSETv0:AI Applications:Facial recognition": 185, - "CSETv1_Annotator-3:Location Country (two letters):US": 185, - "CSETv1_Annotator-1:Location Region:Europe": 183, - "CSETv1_Annotator-2:Date of Incident Year:2016": 183, - "CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:yes": 183, - "CSETv1:Date of Incident Year:2018": 181, - "CSETv1:Tangible Harm:imminent risk of tangible harm (near miss) did occur": 180, - "CSETv1_Annotator-1:Producer Test in Operational Conditions:yes": 180, - "CSETv1_Annotator-2:User Test in Operational Conditions:yes": 180, - "CSETv1_Annotator-3:Physical Objects:yes": 180, - "CSETv1_Annotator-3:Special Interest Intangible Harm:yes": 180, - "CSETv0:Harm Type:Harm to civil liberties": 179, - "CSETv1:Location Region:Asia": 177, - "CSETv0:Nature of End User:Expert": 176, - "CSETv1_Annotator-3:Sector of Deployment:information and communication": 176, - "CSETv1_Annotator-2:Harm Distribution Basis:race": 174, - "CSETv1_Annotator-3:Protected Characteristic:yes": 170, - "CSETv1:Rights Violation:yes": 168, - "CSETv1_Annotator-1:Location Region:Asia": 163, - "CSETv1_Annotator-1:Involving Minor:yes": 157, - "CSETv0:Sector of Deployment:Arts, entertainment and recreation": 156, - "CSETv1_Annotator-3:AI Harm Level:AI tangible harm event": 156, - "CSETv1_Annotator-1:AI Harm Level:AI tangible harm near-miss": 155, - "CSETv1_Annotator-1:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred": 155, - "CSETv0:Sector of Deployment:Public administration and defence": 154, - "CSETv1:Sector of Deployment:wholesale and retail trade": 154, - "CSETv1_Annotator-1:Data Inputs:camera input": 154, - "CSETv1_Annotator-1:Sector of Deployment:wholesale and retail trade": 154, - "CSETv1:Location Region:Europe": 152, - "CSETv1_Annotator-2:Data Inputs:video": 150, - "CSETv1_Annotator-3:Tangible Harm:no tangible harm, near-miss, or issue": 149, - "CSETv0:AI Applications:autonomous driving": 147, - "CSETv1:Data Inputs:sensor data": 147, - "CSETv1_Annotator-2:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred": 147, - "CSETv1_Annotator-1:Date of Incident Year:2023": 145, - "CSETv0:Harm Distribution Basis:Sex": 144, - "CSETv1:Data Inputs:audio": 143, - "CSETv1:Data Inputs:lidar": 143, - "CSETv1_Annotator-1:Entertainment Industry:maybe": 142, - "CSETv0:Public Sector Deployment:true": 141, - "CSETv1_Annotator-3:Date of Incident Year:2017": 141, - "CSETv1:Entertainment Industry:yes": 140, - "CSETv0:AI Techniques:Facial recognition": 139, - "CSETv1:Infrastructure Sectors:transportation": 139, - "CSETv1_Annotator-3:Location Region:Global": 138, - "CSETv0:Named Entities:Amazon": 136, - "CSETv1_Annotator-1:AI tools and methods:natural language processing": 135, - "GMF:Potential AI Technical Failure:Generalization Failure": 135, - "CSETv1_Annotator-1:Data Inputs:facial images": 134, - "CSETv0:AI Techniques:Natural language processing": 133, - "CSETv0:Harm Type:Harm to physical property": 132, - "CSETv1_Annotator-1:AI Harm Level:AI tangible harm issue": 132, - "CSETv1:AI Task:facial recognition": 131, - "CSETv1:Data Inputs:radar": 131, - "CSETv0:Technology Purveyor:Amazon": 129, - "CSETv1:Producer Test in Operational Conditions:yes": 129, - "CSETv1_Annotator-1:Date of Incident Year:2020": 129, - "CSETv1_Annotator-2:AI tools and methods:computer vision": 129, - "CSETv1_Annotator-1:Date of Incident Year:2021": 128, - "CSETv1_Annotator-1:Date of Incident Year:2022": 128, - "CSETv1_Annotator-2:Public Sector Deployment:yes": 128, - "CSETv1_Annotator-2:AI Harm Level:AI tangible harm event": 126, - "CSETv1_Annotator-2:Location Region:Asia": 126, - "CSETv1:Sector of Deployment:administrative and support service activities": 125, - "CSETv1_Annotator-3:Autonomy Level:Autonomy3": 125, - "CSETv1_Annotator-1:Rights Violation:yes": 124, - "CSETv1_Annotator-1:Sector of Deployment:public administration": 123, - "CSETv0:Infrastructure Sectors:Transportation": 122, - "CSETv0:Level of Autonomy:Low": 122, - "CSETv1:Sector of Deployment:public administration": 117, - "CSETv1_Annotator-1:AI Task:facial recognition": 117, - "CSETv1_Annotator-2:Harm Distribution Basis:sex": 116, - "CSETv1_Annotator-1:Harm Distribution Basis:nation of origin, citizenship, immigrant status": 115, - "CSETv1_Annotator-1:Sector of Deployment:professional, scientific and technical activities": 115, - "CSETv0:Severity:Unclear/unknown": 114, - "CSETv1:Sector of Deployment:law enforcement": 114, - "CSETv1_Annotator-1:AI Task:navigation": 114, - "CSETv0:AI Applications:decision support": 113, - "CSETv0:System Developer:Amazon": 112, - "CSETv1:AI Task:navigation": 112, - "CSETv1_Annotator-2:Involving Minor:yes": 111, - "CSETv1_Annotator-2:Detrimental Content:yes": 110, - "CSETv0:Physical System:Consumer device": 108, - "CSETv1_Annotator-1:Clear link to Technology:maybe": 108, - "CSETv1_Annotator-1:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm": 107, - "CSETv1_Annotator-2:Sector of Deployment:transportation and storage": 107, - "CSETv1_Annotator-3:AI tools and methods:unclear": 107, - "CSETv0:AI Applications:recommendation engine": 106, - "CSETv1:Impact on Critical Services:yes": 106, - "CSETv1:Involving Minor:yes": 106, - "CSETv1_Annotator-3:Data Inputs:text": 106, - "CSETv0:Severity:Moderate": 105, - "CSETv1:Clear link to Technology:no": 105, - "CSETv1:AI Harm Level:AI tangible harm near-miss": 104, - "CSETv1:Autonomy Level:unclear": 104, - "CSETv1_Annotator-1:AI tools and methods:facial recognition": 104, - "CSETv1_Annotator-2:Data Inputs:images": 104, - "CSETv0:Lives Lost:true": 103, - "CSETv0:Technology Purveyor:Tesla": 103, - "CSETv1:Deployed:no": 103, - "CSETv1_Annotator-1:AI Task:autonomous navigation": 103, - "CSETv0:Severity:Severe": 102, - "CSETv1_Annotator-1:Autonomy Level:unclear": 102, - "CSETv1_Annotator-1:Date of Incident Year:2014": 102, - "CSETv1_Annotator-2:Sector of Deployment:wholesale and retail trade": 102, - "CSETv1_Annotator-2:Location Region:Europe": 101, - "CSETv1:Data Inputs:facial images": 100, - "CSETv1:Harm Distribution Basis:nation of origin, citizenship, immigrant status": 100, - "CSETv1_Annotator-3:Harm Distribution Basis:race": 100, - "CSETv1_Annotator-1:AI Task:text generation": 99, - "CSETv1_Annotator-2:Sector of Deployment:administrative and support service activities": 99, - "CSETv1:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred": 97, - "CSETv1_Annotator-1:AI Task:semi-autonomous navigation": 96, - "CSETv1:Data Inputs:camera input": 95, - "CSETv1:Date of Incident Year:2014": 95, - "CSETv1_Annotator-2:Date of Incident Year:2015": 95, - "CSETv1_Annotator-2:Rights Violation:yes": 94, - "GMF:Potential AI Technical Failure:Context Misidentification": 94, - "CSETv1:AI System:maybe": 92, - "CSETv1:Sector of Deployment:human health and social work activities": 92, - "CSETv1_Annotator-1:Detrimental Content:maybe": 91, - "CSETv1_Annotator-3:User Test in Operational Conditions:yes": 91, - "CSETv0:AI Applications:image classification": 90, - "CSETv0:Sector of Deployment:Administrative and support service activities": 90, - "CSETv1_Annotator-1:Harm Distribution Basis:religion": 90, - "CSETv1_Annotator-1:Location Region:Oceania": 90, - "CSETv1_Annotator-3:Rights Violation:yes": 90, - "CSETv1_Annotator-2:Data Inputs:audio": 89, - "CSETv0:Level of Autonomy:Unclear/unknown": 88, - "CSETv1:Harm Distribution Basis:religion": 88, - "CSETv1_Annotator-1:AI Task:content generation": 88, - "CSETv1_Annotator-1:AI System:maybe": 86, - "CSETv1_Annotator-1:Data Inputs:radar": 86, - "CSETv0:AI Applications:Natural language processing": 85, - "CSETv0:Intent:Deliberate or expected": 85, - "CSETv1_Annotator-1:Harm Distribution Basis:financial means": 84, - "CSETv1:Date of Incident Year:2020": 83, - "CSETv1_Annotator-2:Report, Test, or Study of data:yes": 83, - "CSETv1:Date of Incident Year:2021": 82, - "CSETv1:Rights Violation:maybe": 82, - "CSETv1_Annotator-2:AI Harm Level:AI tangible harm issue": 82, - "CSETv1_Annotator-3:Clear link to Technology:no": 82, - "CSETv1_Annotator-3:Sector of Deployment:law enforcement": 82, - "CSETv1_Annotator-1:Producer Test in Controlled Conditions:yes": 80, - "CSETv0:Named Entities:Tesla": 79, - "CSETv1_Annotator-1:AI Task:chat bot": 79, - "CSETv1_Annotator-1:Data Inputs:spatial data": 79, - "CSETv1_Annotator-1:Deployed:no": 79, - "CSETv0:AI Applications:image recognition": 78, - "CSETv1:Date of Incident Year:2015": 78, - "CSETv1_Annotator-1:Data Inputs:search queries": 78, - "CSETv1_Annotator-3:Public Sector Deployment:yes": 78, - "CSETv1:AI tools and methods:Natural Language Processesing": 77, - "CSETv1:Date of Incident Year:2023": 77, - "CSETv1_Annotator-1:AI tools and methods:machine learning": 77, - "CSETv1_Annotator-1:Harm Distribution Basis:sexual orientation or gender identity": 77, - "CSETv0:AI Techniques:Unclear": 76, - "CSETv1:Lives Lost:1": 76, - "CSETv1_Annotator-2:Deployed:no": 76, - "CSETv0:AI Techniques:Tesla Autopilot": 75, - "CSETv0:System Developer:Tesla": 75, - "CSETv1:Infrastructure Sectors:healthcare and public health": 75, - "CSETv1_Annotator-1:Deployed:maybe": 75, - "CSETv1_Annotator-2:Date of Incident Year:2018": 74, - "CSETv1_Annotator-2:Infrastructure Sectors:transportation": 74, - "CSETv1:Harm Distribution Basis:sexual orientation or gender identity": 73, - "CSETv1_Annotator-1:Data Inputs:radar input": 73, - "CSETv1_Annotator-2:Autonomy Level:unclear": 73, - "CSETv1:Sector of Deployment:professional, scientific and technical activities": 71, - "CSETv1_Annotator-1:Date of Incident Year:2015": 71, - "CSETv1_Annotator-1:Sector of Deployment:financial and insurance activities": 71, - "CSETv1:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm": 70, - "CSETv1_Annotator-3:Sector of Deployment:transportation and storage": 70, - "CSETv0:Relevant AI functions:Unclear": 69, - "CSETv1:Producer Test in Controlled Conditions:yes": 69, - "CSETv1:Sector of Deployment:financial and insurance activities": 69, - "GMF:Known AI Technology:Face Detection": 69, - "CSETv1:Data Inputs:search queries": 68, - "CSETv1_Annotator-1:Data Inputs:voice input": 68, - "CSETv1_Annotator-1:Sector of Deployment:human health and social work activities": 68, - "CSETv1_Annotator-2:Infrastructure Sectors:healthcare and public health": 68, - "CSETv1_Annotator-2:Sector of Deployment:human health and social work activities": 68, - "CSETv1:Physical System Type:Manufacturing Robot": 67, - "CSETv1:Sector of Deployment:manufacturing": 67, - "CSETv1_Annotator-2:Sector of Deployment:manufacturing": 67, - "CSETv1:AI Harm Level:AI tangible harm issue": 66, - "CSETv1:Injuries:1": 66, - "CSETv1_Annotator-1:Data Inputs:user queries": 66, - "CSETv1_Annotator-1:Lives Lost:1": 66, - "CSETv1_Annotator-1:Location Country (two letters):AU": 66, - "CSETv1_Annotator-2:Producer Test in Controlled Conditions:yes": 66, - "CSETv1_Annotator-3:Data Inputs:video": 65, - "CSETv0:Named Entities:Microsoft": 64, - "CSETv1:Data Inputs:speech": 64, - "CSETv1:Location Region:Oceania": 64, - "CSETv1_Annotator-1:AI Task:self-driving": 64, - "CSETv1_Annotator-1:Data Inputs:geospatial data": 64, - "CSETv1_Annotator-2:AI System:maybe": 64, - "CSETv1_Annotator-2:Date of Incident Year:2023": 64, - "CSETv1_Annotator-2:Sector of Deployment:financial and insurance activities": 64, - "CSETv1:Operating Conditions:operationally representative": 63, - "CSETv1:Report, Test, or Study of data:yes": 63, - "CSETv0:AI Applications:self-driving vehicle": 62, - "CSETv0:AI Techniques:Deep learning": 62, - "CSETv0:Named Entities:Knightscope": 62, - "CSETv0:Named Entities:Knightscope K5": 62, - "CSETv0:System Developer:Knightscope": 62, - "CSETv0:Technology Purveyor:Knightscope": 62, - "CSETv1:AI Task:patrolling": 62, - "CSETv1_Annotator-1:AI Task:virtual assistant technology": 62, - "CSETv1_Annotator-1:Physical System Type:rocket/egg shaped, 300 pound, 5 ft. tall security robot": 62, - "CSETv1_Annotator-2:Data Inputs:lidar": 62, - "CSETv1_Annotator-2:Location Region:Oceania": 62, - "CSETv1_Annotator-3:Location Region:Asia": 62, - "CSETv1:Data Inputs:video input": 61, - "CSETv1:Data Inputs:voice": 61, - "CSETv1_Annotator-1:Injuries:1": 61, - "CSETv1_Annotator-2:Data Inputs:voice": 61, - "CSETv0:Sector of Deployment:Financial and insurance activities": 60, - "CSETv1_Annotator-1:Data Inputs:lidar": 60, - "CSETv1_Annotator-1:Data Inputs:user prompts": 60, - "CSETv1_Annotator-2:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm": 60, - "CSETv1_Annotator-2:Sector of Deployment:law enforcement": 60, - "CSETv0:Harm Distribution Basis:National origin or immigrant status": 59, - "CSETv1:Location Country (two letters):RU": 59, - "CSETv1:Physical System Type:none": 59, - "CSETv1_Annotator-1:AI tools and methods:voice recognition": 59, - "CSETv1_Annotator-1:Data Inputs:speech": 59, - "CSETv1_Annotator-1:Report, Test, or Study of data:maybe": 59, - "CSETv1_Annotator-2:Physical System Type:none": 59, - "GMF:Known AI Technical Failure:Unsafe Exposure or Access": 59, - "CSETv1:AI Task:virtual assistant technology": 58, - "CSETv1:Data Inputs:photos": 58, - "CSETv0:AI Techniques:environmental sensing": 57, - "CSETv0:Named Entities:Stanford Shopping Center": 57, - "CSETv1:AI Task:security": 57, - "CSETv1:Data Inputs:road data": 57, - "CSETv1:Physical System Type:rocket/egg shaped, 300 pound, 5 ft. tall security robot": 57, - "CSETv1_Annotator-1:AI Task:patrolling": 57, - "CSETv1_Annotator-1:AI Task:security": 57, - "CSETv1_Annotator-2:AI tools and methods:Text to Speach": 57, - "CSETv1_Annotator-2:Data Inputs:road data": 57, - "CSETv1_Annotator-2:Operating Conditions:operationally representative": 57, - "CSETv1_Annotator-1:Location Country (two letters):CN": 56, - "CSETv1_Annotator-3:Date of Incident Year:2016": 56, - "GMF:Potential AI Technical Failure:Dataset Imbalance": 56, - "CSETv0:AI Techniques:open-source": 55, - "CSETv0:Technology Purveyor:Microsoft": 55, - "CSETv1:AI Task:object detection": 55, - "CSETv1:AI tools and methods:facial recognition": 55, - "CSETv1:AI tools and methods:voice recognition": 55, - "CSETv1_Annotator-1:Physical Objects:maybe": 55, - "CSETv1_Annotator-1:Report, Test, or Study of data:yes": 55, - "CSETv1_Annotator-1:Rights Violation:maybe": 55, - "CSETv1_Annotator-1:Sector of Deployment:manufacturing": 55, - "CSETv1_Annotator-2:AI tools and methods:Audio Transcription": 55, - "CSETv1_Annotator-2:Physical System Type:Manufacturing Robot": 55, - "GMF:Known AI Technical Failure:Distributional Bias": 55, - "GMF:Known AI Technology:Recurrent Neural Network": 55, - "CSETv1:AI Task:autonomous driving": 54, - "CSETv1:AI tools and methods:neural networks": 54, - "CSETv1_Annotator-1:Data Inputs:audio inputs": 54, - "GMF:Potential AI Technology:Regression": 54, - "CSETv0:Harm Distribution Basis:Financial means": 53, - "CSETv1:AI Task:self-driving": 53, - "CSETv1:Date of Incident Year:2013": 53, - "CSETv1:Location Country (two letters):": 53, - "CSETv1_Annotator-1:AI Task:image identification": 53, - "CSETv1_Annotator-2:Infrastructure Sectors:communications": 53, - "GMF:Known AI Technology:Distributional Learning": 53, - "CSETv0:AI Techniques:image classification": 52, - "CSETv1:Clear link to Technology:maybe": 52, - "CSETv1:Data Inputs:spatial data": 52, - "CSETv1_Annotator-1:Date of Incident Year:2019": 52, - "CSETv1_Annotator-1:Sector of Deployment:Education": 52, - "GMF:Potential AI Technical Failure:Underspecification": 52, - "CSETv0:Location:Palo Alto, CA": 51, - "CSETv0:Named Entities:Apple": 51, - "CSETv0:Named Entities:Joshua Brown": 51, - "CSETv0:Technology Purveyor:Apple": 51, - "CSETv1:AI Task:object recognition": 51, - "CSETv1:AI tools and methods:audio transcription": 51, - "CSETv1:AI tools and methods:machine learning": 51, - "CSETv1:Data Inputs:infrared camera": 51, - "CSETv1:Data Inputs:personal data": 51, - "CSETv1_Annotator-1:Data Inputs:infrared camera": 51, - "CSETv1_Annotator-2:Impact on Critical Services:yes": 51, - "CSETv1_Annotator-2:Physical System Type:Amazon Alexa": 51, - "CSETv1_Annotator-1:Special Interest Intangible Harm:maybe": 50, - "GMF:Known AI Technology:Convolutional Neural Network": 50, - "CSETv0:AI Applications:chatbot": 49, - "CSETv0:Harm Distribution Basis:Religion": 49, - "CSETv1:AI Task:chat bot": 49, - "CSETv1_Annotator-1:AI Task:smart suggestions": 49, - "CSETv1:Multiple AI Interaction:yes": 48, - "CSETv1_Annotator-1:AI Harm Level:unclear": 48, - "CSETv1_Annotator-1:Date of Incident Year:2013": 48, - "CSETv1_Annotator-1:Infrastructure Sectors:transportation": 48, - "CSETv1_Annotator-1:Multiple AI Interaction:yes": 48, - "CSETv1_Annotator-3:AI Harm Level:unclear": 48, - "CSETv1_Annotator-3:Data Inputs:audio": 48, - "CSETv1_Annotator-3:Location Region:Europe": 48, - "CSETv1:Data Inputs:image alt tags": 47, - "CSETv1_Annotator-1:Data Inputs:image alt tags": 47, - "CSETv1_Annotator-2:AI Task:autonomous driving": 47, - "CSETv1_Annotator-3:Impact on Critical Services:yes": 47, - "CSETv0:AI Techniques:autonomous vehicle": 46, - "CSETv1:AI Task:text generation": 46, - "CSETv1_Annotator-1:Impact on Critical Services:yes": 46, - "CSETv1_Annotator-2:Location Country (two letters):": 46, - "CSETv0:System Developer:Apple": 45, - "CSETv1:AI tools and methods:image mapping": 45, - "CSETv1:AI tools and methods:point mapping": 45, - "CSETv1:Data Inputs:dot projector": 45, - "CSETv1:Location Country (two letters):CN": 45, - "CSETv1:Physical System Type:Apple iPhone X": 45, - "CSETv1_Annotator-1:AI Task:image generation": 45, - "CSETv1_Annotator-1:AI tools and methods:facial reconstruction": 45, - "CSETv1_Annotator-1:AI tools and methods:image mapping": 45, - "CSETv1_Annotator-1:AI tools and methods:image reconstruction": 45, - "CSETv1_Annotator-1:AI tools and methods:point mapping": 45, - "CSETv1_Annotator-1:Data Inputs:dot projector": 45, - "CSETv1_Annotator-1:Physical System Type:Apple iPhone X": 45, - "CSETv1_Annotator-2:Date of Incident Year:2020": 45, - "CSETv1_Annotator-2:Producer Test in Operational Conditions:yes": 45, - "CSETv1_Annotator-3:AI System:no": 45, - "CSETv1_Annotator-3:Data Inputs:radar": 45, - "CSETv1_Annotator-3:Harm Distribution Basis:sex": 45, - "CSETv1_Annotator-3:Harm Distribution Basis:sexual orientation or gender identity": 45, - "GMF:Known AI Goal:Autonomous Driving": 45, - "GMF:Potential AI Technical Failure:Unauthorized Data": 45, - "GMF:Potential AI Technology:Image Segmentation": 45, - "CSETv0:AI Applications:voice recognition": 44, - "CSETv0:System Developer:Microsoft": 44, - "CSETv1:AI Task:search optimization": 44, - "CSETv1:Data Inputs:GPS": 44, - "CSETv1:Harm Distribution Basis:financial means": 44, - "CSETv1:Intentional Harm:unclear": 44, - "CSETv1_Annotator-2:Entertainment Industry:yes": 44, - "CSETv1_Annotator-3:Harm Distribution Basis:religion": 44, - "GMF:Potential AI Technology:Classification": 44, - "GMF:Potential AI Technology:Convolutional Neural Network": 44, - "CSETv0:AI Techniques:language recognition": 43, - "CSETv1:AI Task:autonomous navigation": 43, - "CSETv1:Date of Incident Year:2011": 43, - "CSETv1:Date of Incident Year:2019": 43, - "CSETv1_Annotator-1:AI Task:content moderation": 43, - "CSETv1_Annotator-1:AI Task:search optimization": 43, - "CSETv1_Annotator-1:AI Task:search suggestion": 43, - "CSETv1_Annotator-1:Location Country (two letters):DE": 43, - "CSETv1_Annotator-2:AI Task:self driving": 43, - "CSETv1_Annotator-2:Multiple AI Interaction:yes": 43, - "GMF:Known AI Technology:Language Modeling": 43, - "CSETv0:Location:Los Angeles, CA": 42, - "CSETv1:AI tools and methods:unclear": 42, - "CSETv1_Annotator-1:Date of Incident Year:2012": 42, - "CSETv1_Annotator-1:Sector of Deployment:administrative and support service activities": 42, - "CSETv1_Annotator-3:Detrimental Content:yes": 42, - "CSETv0:Named Entities:China": 41, - "CSETv1:AI Task:search suggestion": 41, - "CSETv1:Date of Incident Year:2012": 41, - "CSETv1:Entertainment Industry:maybe": 41, - "CSETv1:Harm Distribution Basis:disability": 41, - "CSETv1:Location Country (two letters):IN": 41, - "CSETv1:Sector of Deployment:Education": 41, - "CSETv1_Annotator-1:Data Inputs:age": 41, - "CSETv1_Annotator-1:Harm Distribution Basis:disability": 41, - "CSETv1_Annotator-1:Lives Lost:2": 41, - "CSETv1_Annotator-2:Date of Incident Year:2014": 41, - "CSETv1_Annotator-2:Location Country (two letters):IN": 41, - "CSETv1_Annotator-3:AI System:maybe": 41, - "CSETv1_Annotator-3:Autonomy Level:Autonomy2": 41, - "GMF:Potential AI Technical Failure:Underfitting": 41, - "CSETv0:AI Applications:environmental sensing": 40, - "CSETv0:AI Techniques:virtual assistant": 40, - "CSETv0:Harm Distribution Basis:Sexual orientation or gender identity": 40, - "CSETv1:AI Task:Image search": 40, - "CSETv1:Location Country (two letters):AU": 40, - "CSETv1_Annotator-1:Harm Distribution Basis:ideology": 40, - "CSETv1_Annotator-2:AI Task:Personal Assistant": 40, - "CSETv1_Annotator-2:Date of Incident Year:2021": 40, - "CSETv1_Annotator-3:Tangible Harm:imminent risk of tangible harm (near miss) did occur": 40, - "CSETv0:AI Techniques:Amazon Alexa": 39, - "CSETv0:Location:Australia": 39, - "CSETv0:Named Entities:Alan Tudge": 39, - "CSETv0:Named Entities:Centrelink": 39, - "CSETv0:Named Entities:Centrelink Master Program": 39, - "CSETv0:Named Entities:Department of Human Services": 39, - "CSETv0:Named Entities:Richard Glenn": 39, - "CSETv0:System Developer:Centrelink Master Program": 39, - "CSETv0:Technology Purveyor:Centrelink Master Program": 39, - "CSETv0:Technology Purveyor:Department of Human Services": 39, - "CSETv1:Data Inputs:income data": 39, - "CSETv1:Data Inputs:tabular": 39, - "CSETv1:Data Inputs:tax reports": 39, - "CSETv1:Detrimental Content:maybe": 39, - "CSETv1_Annotator-1:Data Inputs:income data": 39, - "CSETv1_Annotator-1:Data Inputs:laser input": 39, - "CSETv1_Annotator-1:Data Inputs:tax reports": 39, - "CSETv1_Annotator-1:Data Inputs:words": 39, - "CSETv1_Annotator-1:Infrastructure Sectors:government facilities": 39, - "CSETv1_Annotator-2:Data Inputs:amazon account": 39, - "CSETv1_Annotator-2:Infrastructure Sectors:government facilities": 39, - "CSETv1_Annotator-2:Lives Lost:1": 39, - "CSETv1_Annotator-2:Location Country (two letters):AU": 39, - "CSETv1_Annotator-3:Autonomy Level:unclear": 39, - "CSETv1_Annotator-3:Data Inputs:lidar": 39, - "CSETv1_Annotator-3:Sector of Deployment:wholesale and retail trade": 39, - "GMF:Potential AI Technical Failure:Hardware Failure": 39, - "GMF:Potential AI Technical Failure:Lack of Capability Control": 39, - "GMF:Potential AI Technology:Ensemble Aggregation": 39, - "CSETv0:AI Techniques:natural language processing model": 38, - "CSETv1:AI Task:predict recidivism": 38, - "CSETv1:AI tools and methods:object detection": 38, - "CSETv1_Annotator-1:AI Task:predict recidivism": 38, - "CSETv1_Annotator-1:AI tools and methods:computer vision": 38, - "CSETv1_Annotator-2:AI tools and methods:Machine Learning": 38, - "CSETv1_Annotator-2:Sector of Deployment:other": 38, - "GMF:Potential AI Technical Failure:Data or Labelling Noise": 38, - "CSETv0:AI Techniques:law enforcement algorithm": 37, - "CSETv0:Infrastructure Sectors:Government facilities": 37, - "CSETv1:AI Task:Rank Applicants": 37, - "CSETv1:AI Task:translation": 37, - "CSETv1:Data Inputs:Amazon account information": 37, - "CSETv1:Deployed:maybe": 37, - "CSETv1_Annotator-1:AI Task:search result ranking": 37, - "CSETv1_Annotator-1:Data Inputs:137 questionnaire responses": 37, - "CSETv1_Annotator-1:Data Inputs:Amazon account information": 37, - "CSETv1_Annotator-2:AI Task:Rank Applicants": 37, - "CSETv1_Annotator-3:Date of Incident Year:2020": 37, - "GMF:Known AI Technical Failure:Misuse": 37, - "GMF:Known AI Technology:Automatic Speech Recognition": 37, - "CSETv0:AI Applications:language translation": 36, - "CSETv1:AI Harm Level:unclear": 36, - "CSETv1:Data Inputs:Facebook posts": 36, - "CSETv1:Data Inputs:air quality": 36, - "CSETv1:Date of Incident Year:2022": 36, - "CSETv1:Infrastructure Sectors:financial services": 36, - "CSETv1_Annotator-1:AI Task:transportation": 36, - "CSETv1_Annotator-1:Location Country (two letters):IN": 36, - "CSETv1_Annotator-1:Location Country (two letters):RU": 36, - "CSETv1_Annotator-1:Physical System Type:Tesla Model S": 36, - "CSETv1_Annotator-3:Entertainment Industry:yes": 36, - "GMF:Known AI Goal:Face Recognition": 36, - "CSETv0:AI Applications:Interpreting traffic patterns": 35, - "CSETv0:AI Applications:image processing": 35, - "CSETv0:Harm Distribution Basis:Geography": 35, - "CSETv0:Infrastructure Sectors:Information technology": 35, - "CSETv0:Named Entities:San Diego TV": 35, - "CSETv1:AI Task:prediction": 35, - "CSETv1:AI Task:self driving": 35, - "CSETv1:AI Task:semi-autonomous navigation": 35, - "CSETv1:AI Task:surveillance": 35, - "CSETv1:Data Inputs:Crime Records": 35, - "CSETv1:Harm Distribution Basis:geography": 35, - "CSETv1:Location Country (two letters):IE": 35, - "CSETv1:Physical System Type:Amazon Alexa Echo Dot speaker": 35, - "CSETv1:Physical System Type:Volvo XC90 SUV": 35, - "CSETv1_Annotator-1:AI Task:image modification": 35, - "CSETv1_Annotator-1:AI tools and methods:audio transcription": 35, - "CSETv1_Annotator-1:Location Country (two letters):IE": 35, - "CSETv1_Annotator-1:Physical System Type:Amazon Alexa Echo Dot speaker": 35, - "CSETv1_Annotator-1:Sector of Deployment:defense": 35, - "CSETv1_Annotator-2:AI Harm Level:AI tangible harm near-miss": 35, - "CSETv1_Annotator-2:Deployed:maybe": 35, - "CSETv1_Annotator-3:Data Inputs:face image": 35, - "CSETv1_Annotator-3:Date of Incident Year:2021": 35, - "GMF:Known AI Goal:AI Voice Assistant": 35, - "GMF:Known AI Technology:Acoustic Fingerprint": 35, - "GMF:Known AI Technology:Neural Network": 35, - "GMF:Potential AI Technical Failure:Inadequate Anonymization": 35, - "GMF:Potential AI Technology:Distributional Learning": 35, - "CSETv0:AI Techniques:radar": 34, - "CSETv0:Harm Type:Harm to intangible property": 34, - "CSETv0:Sector of Deployment:Professional, scientific and technical activities": 34, - "CSETv1:AI Task:Driving": 34, - "CSETv1_Annotator-1:Data Inputs:photos": 34, - "CSETv1_Annotator-1:Protected Characteristic:maybe": 34, - "CSETv1_Annotator-1:Public Sector Deployment:maybe": 34, - "CSETv1_Annotator-2:AI Task:recommender": 34, - "CSETv1_Annotator-2:Sector of Deployment:Arts, entertainment and recreation": 34, - "GMF:Potential AI Technical Failure:Overfitting": 34, - "CSETv0:Location:Edinburgh, Scotland": 33, - "CSETv0:Named Entities:Edinburgh": 33, - "CSETv1:AI Task:resume screening": 33, - "CSETv1:Data Inputs:Resume": 33, - "CSETv1:Data Inputs:traffic": 33, - "CSETv1:Harm Distribution Basis:ideology": 33, - "CSETv1_Annotator-1:AI Task:obstacle avoidance": 33, - "CSETv1_Annotator-1:AI Task:recruitment": 33, - "CSETv1_Annotator-1:AI Task:resume screening": 33, - "CSETv1_Annotator-1:Data Inputs:resumes": 33, - "CSETv1_Annotator-2:AI tools and methods:Natural language processing": 33, - "CSETv1_Annotator-2:Data Inputs:Resume": 33, - "CSETv1_Annotator-2:Data Inputs:personal data": 33, - "CSETv1_Annotator-2:Data Inputs:photos": 33, - "CSETv1_Annotator-2:Data Inputs:traffic": 33, - "CSETv1_Annotator-2:Date of Incident Year:2013": 33, - "CSETv1_Annotator-2:Infrastructure Sectors:unclear": 33, - "CSETv1_Annotator-2:Location Country (two letters):IE": 33, - "CSETv1_Annotator-2:Rights Violation:maybe": 33, - "CSETv1_Annotator-3:Tangible Harm:unclear": 33, - "GMF:Known AI Technology:Generative Adversarial Network": 33, - "GMF:Potential AI Technical Failure:Misinformation Generation Hazard": 33, - "CSETv0:AI Applications:content filtering": 32, - "CSETv0:AI Applications:environment sensing": 32, - "CSETv0:AI Applications:self-driving": 32, - "CSETv0:Named Entities:Facebook": 32, - "CSETv0:System Developer:Facebook": 32, - "CSETv0:Technology Purveyor:Facebook": 32, - "CSETv1:AI tools and methods:human language technology": 32, - "CSETv1:Location Country (two letters):DE": 32, - "CSETv1_Annotator-1:AI Task:deepfake video generation": 32, - "CSETv1:AI Task:content generation": 31, - "CSETv1:AI Task:deepfake video generation": 31, - "CSETv1:AI tools and methods:search engine optimization": 31, - "CSETv1:Data Inputs:age": 31, - "CSETv1:Data Inputs:personal information": 31, - "CSETv1:Infrastructure Sectors:defense-industrial base": 31, - "CSETv1:Sector of Deployment:defense": 31, - "CSETv1_Annotator-1:AI Task:image search": 31, - "CSETv1_Annotator-1:AI tools and methods:neural networks": 31, - "CSETv1_Annotator-1:AI tools and methods:search engine optimization": 31, - "CSETv1_Annotator-1:Sector of Deployment:other": 31, - "CSETv1_Annotator-2:Location Country (two letters):DE": 31, - "GMF:Known AI Goal:Autonomous Drones": 31, - "CSETv0:AI Applications:forecasting": 30, - "CSETv0:AI Applications:stock trading": 30, - "CSETv0:AI Techniques:stock market algorithm": 30, - "CSETv0:Financial Cost:Short term: $1 trillion unclear of long-term impact": 30, - "CSETv0:Infrastructure Sectors:Financial services": 30, - "CSETv0:Location:UK/USA": 30, - "CSETv0:Location:Washington, D.C.": 30, - "CSETv0:Named Entities:Chicago Merchant Exchange": 30, - "CSETv0:Named Entities:Dow Jones Industrial Index": 30, - "CSETv0:Named Entities:Electronic Privacy Information Center (EPIC)": 30, - "CSETv0:Named Entities:Mountain View": 30, - "CSETv0:Named Entities:Navinder Singh Saroa": 30, - "CSETv0:Technology Purveyor:Navinder Singh Sarao": 30, - "CSETv1:AI Task:autocomplete": 30, - "CSETv1:AI Task:content moderation": 30, - "CSETv1:AI Task:recommender": 30, - "CSETv1:Data Inputs:radar input": 30, - "CSETv1:Data Inputs:stock data": 30, - "CSETv1:Data Inputs:stock orders": 30, - "CSETv1:Data Inputs:stock prices": 30, - "CSETv1:Data Inputs:thermal imaging": 30, - "CSETv1:Data Inputs:words": 30, - "CSETv1:Date of Incident Year:2010": 30, - "CSETv1:Harm Distribution Basis:age": 30, - "CSETv1_Annotator-1:AI Task:surveillance": 30, - "CSETv1_Annotator-1:Data Inputs:Youtube videos": 30, - "CSETv1_Annotator-1:Data Inputs:air quality data": 30, - "CSETv1_Annotator-1:Data Inputs:personal information": 30, - "CSETv1_Annotator-1:Data Inputs:stock orders": 30, - "CSETv1_Annotator-1:Data Inputs:stock prices": 30, - "CSETv1_Annotator-1:Data Inputs:thermal imaging": 30, - "CSETv1_Annotator-1:Date of Incident Year:2010": 30, - "CSETv1_Annotator-1:Infrastructure Sectors:financial services": 30, - "CSETv1_Annotator-2:Data Inputs:stock data": 30, - "CSETv1_Annotator-2:Date of Incident Year:2010": 30, - "CSETv1_Annotator-3:AI Task:Patrolling": 30, - "CSETv1_Annotator-3:Data Inputs:Facebook posts": 30, - "CSETv1_Annotator-3:Data Inputs:Unclear": 30, - "CSETv1_Annotator-3:Data Inputs:air quality": 30, - "CSETv1_Annotator-3:Data Inputs:thermal imaging": 30, - "CSETv1_Annotator-3:Physical System Type:autonomous security robot": 30, - "GMF:Known AI Goal:Automatic Stock Trading": 30, - "GMF:Potential AI Technical Failure:Gaming Vulnerability": 30, - "CSETv0:AI Applications:Content curation": 29, - "CSETv0:AI Techniques:GANs": 29, - "CSETv0:AI Techniques:RNNs": 29, - "CSETv0:Harm Type:Other:Reputational harm/social harm (libel and defamation)": 29, - "CSETv0:Location:Delhi, India": 29, - "CSETv0:Location:Unknown": 29, - "CSETv0:Location:Williston, FL": 29, - "CSETv0:Named Entities:Barack Obama": 29, - "CSETv0:Named Entities:Delhi Metro Rail Corporation": 29, - "CSETv0:Named Entities:Hyuandai Rotem": 29, - "CSETv0:Named Entities:Jordan Peele": 29, - "CSETv0:Named Entities:Tesla Autopilot": 29, - "CSETv0:Named Entities:Tesla Model S": 29, - "CSETv0:Named Entities:University of Washington": 29, - "CSETv0:System Developer:FakeApp": 29, - "CSETv0:System Developer:University of Washington": 29, - "CSETv0:Technology Purveyor:Delhi Metro Rail Corporation": 29, - "CSETv0:Technology Purveyor:Hyuandai Rotem": 29, - "CSETv0:Technology Purveyor:Jordan Peele": 29, - "CSETv0:Technology Purveyor:University of Washington": 29, - "CSETv1:AI Task:content ranking": 29, - "CSETv1:AI tools and methods:Sensor Data Processing": 29, - "CSETv1:Data Inputs:Train position/status": 29, - "CSETv1:Data Inputs:geospatial data": 29, - "CSETv1:Data Inputs:track sensors": 29, - "CSETv1:Infrastructure Sectors:communications": 29, - "CSETv1:Operating Conditions:Testing": 29, - "CSETv1:Physical System Type:Camera": 29, - "CSETv1:Physical System Type:Metro train": 29, - "CSETv1_Annotator-1:AI Task:autocomplete": 29, - "CSETv1_Annotator-1:AI Task:object detection": 29, - "CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:maybe": 29, - "CSETv1_Annotator-1:Date of Incident Year:2011": 29, - "CSETv1_Annotator-1:Location Region:--": 29, - "CSETv1_Annotator-1:Physical System Type:Automobile": 29, - "CSETv1_Annotator-1:Physical System Type:Metro train": 29, - "CSETv1_Annotator-2:AI Task:Deepfake": 29, - "CSETv1_Annotator-2:Data Inputs:Train position/status": 29, - "CSETv1_Annotator-2:Data Inputs:track sensors": 29, - "CSETv1_Annotator-2:Detrimental Content:maybe": 29, - "CSETv1_Annotator-2:Operating Conditions:Testing": 29, - "CSETv1_Annotator-2:Physical System Type:Metro train": 29, - "CSETv1_Annotator-3:AI Task:autocomplete": 29, - "CSETv1_Annotator-3:AI Task:content ranking": 29, - "CSETv1_Annotator-3:Date of Incident Year:2011": 29, - "GMF:Known AI Goal:Deepfake Video Generation": 29, - "GMF:Known AI Goal:Hate Speech Detection": 29, - "GMF:Known AI Technical Failure:Adversarial Data": 29, - "GMF:Potential AI Technical Failure:Distributional Bias": 29, - "GMF:Potential AI Technical Failure:Misuse": 29, - "GMF:Potential AI Technology:3D reconstruction": 29, - "GMF:Potential AI Technology:Other domain-specific approaches": 29, - "CSETv0:AI Applications:Text generation": 28, - "CSETv0:AI Applications:comprehension": 28, - "CSETv0:AI Applications:computer vision": 28, - "CSETv0:AI Applications:language output": 28, - "CSETv0:AI Techniques:content creation": 28, - "CSETv0:AI Techniques:driving lane detection": 28, - "CSETv0:AI Techniques:image recognition": 28, - "CSETv0:AI Techniques:language recognition natural language processing": 28, - "CSETv0:AI Techniques:object detection": 28, - "CSETv0:AI Techniques:scene segmentation": 28, - "CSETv0:Location:Cambridge, MA": 28, - "CSETv0:Location:Mountain View, CA": 28, - "CSETv0:Location:Storey County, Nevada": 28, - "CSETv0:Named Entities:Lexus": 28, - "CSETv0:Named Entities:MIT Media Lab": 28, - "CSETv0:Named Entities:Massachusetts Intstitute of Technology": 28, - "CSETv0:Named Entities:Norman": 28, - "CSETv0:Named Entities:Panasonic": 28, - "CSETv0:Named Entities:Reddit": 28, - "CSETv0:Named Entities:Santa Clara Valley Transportation Authority": 28, - "CSETv0:Named Entities:Tay": 28, - "CSETv0:Named Entities:Twitter": 28, - "CSETv0:Named Entities:Xiaoice": 28, - "CSETv0:System Developer:MIT Media Lab": 28, - "CSETv0:Technology Purveyor:MIT Media Lab": 28, - "CSETv0:Technology Purveyor:Twitter": 28, - "CSETv1:AI Task:Generate Captions": 28, - "CSETv1:AI Task:assembly": 28, - "CSETv1:AI Task:image identification": 28, - "CSETv1:AI Task:image interpretation": 28, - "CSETv1:AI Task:production": 28, - "CSETv1:AI tools and methods:prediction": 28, - "CSETv1:Data Inputs:Captions": 28, - "CSETv1:Data Inputs:image captions": 28, - "CSETv1:Data Inputs:laser input": 28, - "CSETv1:Data Inputs:video captions": 28, - "CSETv1:Date of Incident Year:2009": 28, - "CSETv1:Operating Conditions:Driverless car": 28, - "CSETv1:Physical System Type:Vehicles (Lexus, Audi, Chrysler Pacifica)": 28, - "CSETv1_Annotator-1:AI Task:assembly": 28, - "CSETv1_Annotator-1:AI Task:image classification": 28, - "CSETv1_Annotator-1:AI Task:image interpretation": 28, - "CSETv1_Annotator-1:AI Task:production": 28, - "CSETv1_Annotator-1:AI tools and methods:human language technology": 28, - "CSETv1_Annotator-1:AI tools and methods:large language models": 28, - "CSETv1_Annotator-1:Data Inputs:image captions": 28, - "CSETv1_Annotator-1:Data Inputs:video captions": 28, - "CSETv1_Annotator-1:Date of Incident Year:2024": 28, - "CSETv1_Annotator-1:Physical System Type:Vehicles (Lexus, Audi, Chrysler Pacifica)": 28, - "CSETv1_Annotator-2:AI Task:Generate Captions": 28, - "CSETv1_Annotator-2:AI Task:Manufacturing Car": 28, - "CSETv1_Annotator-2:Data Inputs:Captions": 28, - "CSETv1_Annotator-2:Harm Distribution Basis:nation of origin, citizenship, immigrant status": 28, - "CSETv1_Annotator-2:Harm Distribution Basis:religion": 28, - "GMF:Known AI Technical Failure:Generalization Failure": 28, - "GMF:Known AI Technical Failure:Limited Dataset": 28, - "CSETv0:AI Applications:Early warning system": 27, - "CSETv0:AI Applications:targeted advertising": 27, - "CSETv0:AI Techniques:Google Adsense": 27, - "CSETv0:AI Techniques:Oko satellites": 27, - "CSETv0:Harm Type:Other:Reputational harm": 27, - "CSETv0:Infrastructure Sectors:Critical manufacturing": 27, - "CSETv0:Infrastructure Sectors:Nuclear": 27, - "CSETv0:Location:Baunatal, Germany": 27, - "CSETv0:Location:Soviet Union": 27, - "CSETv0:Named Entities:Carnegie Mellon University": 27, - "CSETv0:Named Entities:Harvard University": 27, - "CSETv0:Named Entities:Harwin Cheng": 27, - "CSETv0:Named Entities:Oko": 27, - "CSETv0:Named Entities:Soviet Union": 27, - "CSETv0:Named Entities:Tiffany Teng": 27, - "CSETv0:Named Entities:United States": 27, - "CSETv0:Named Entities:Volkswagen": 27, - "CSETv0:Named Entities:William Santana Li": 27, - "CSETv0:Named Entities:www.instantcheckmate.com": 27, - "CSETv0:Physical System:Weapons system": 27, - "CSETv0:System Developer:Soviet Union": 27, - "CSETv0:Technology Purveyor:Instant Checkmate": 27, - "CSETv0:Technology Purveyor:Soviet Union": 27, - "CSETv0:Technology Purveyor:United States": 27, - "CSETv0:Technology Purveyor:Volkswagen": 27, - "CSETv1:AI Task:personalized online advertising": 27, - "CSETv1:Data Inputs:satellite data": 27, - "CSETv1:Data Inputs:sonar": 27, - "CSETv1:Data Inputs:vibration detectors": 27, - "CSETv1:Date of Incident Year:1983": 27, - "CSETv1:Injuries:10": 27, - "CSETv1_Annotator-1:AI Task:personalized online advertising": 27, - "CSETv1_Annotator-1:Data Inputs:satellite data": 27, - "CSETv1_Annotator-1:Date of Incident Year:1983": 27, - "CSETv1_Annotator-1:Injuries:10": 27, - "CSETv1_Annotator-1:Sector of Deployment:accommodation and food service activities": 27, - "CSETv1_Annotator-2:AI Task:facial recognition": 27, - "CSETv1_Annotator-2:AI tools and methods:Sensor Data Processing": 27, - "CSETv1_Annotator-2:Data Inputs:sonar": 27, - "CSETv1_Annotator-2:Data Inputs:vibration detectors": 27, - "CSETv1_Annotator-2:Injuries:1": 27, - "CSETv1_Annotator-2:Location Country (two letters):CN": 27, - "CSETv1_Annotator-2:Physical System Type:5' tall, 300 pound, egg shaped security robot": 27, - "CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:maybe": 27, - "CSETv1_Annotator-3:Clear link to Technology:maybe": 27, - "CSETv1_Annotator-3:Date of Incident Year:1983": 27, - "CSETv1_Annotator-3:Infrastructure Sectors:defense-industrial base": 27, - "CSETv1_Annotator-3:Intentional Harm:Yes. Intentionally designed to perform harm but created an unintended harm (a different harm may have occurred)": 27, - "CSETv1_Annotator-3:Location Country (two letters):RU": 27, - "CSETv1_Annotator-3:Sector of Deployment:defense": 27, - "CSETv1_Annotator-3:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred": 27, - "GMF:Known AI Goal:Threat Detection": 27, - "GMF:Known AI Technical Failure:Black Swan Event": 27, - "GMF:Known AI Technical Failure:Data or Labelling Noise": 27, - "GMF:Known AI Technical Failure:Dataset Imbalance": 27, - "GMF:Potential AI Technology:Face Detection": 27, - "GMF:Potential AI Technology:Satellite Imaging": 27, - "CSETv0:Location:Beitar Illit, Israel": 26, - "CSETv0:Named Entities:Beitar Illit": 26, - "CSETv0:Named Entities:Israel": 26, - "CSETv0:Named Entities:Israeli Police": 26, - "CSETv0:Sector of Deployment:Human health and social work activities": 26, - "CSETv1:Data Inputs:Youtube videos": 26, - "CSETv1:Data Inputs:traffic patterns": 26, - "CSETv1:Location Country (two letters):PS": 26, - "CSETv1:Location Country (two letters):VN": 26, - "CSETv1_Annotator-1:AI Task:route optimization": 26, - "CSETv1_Annotator-1:AI tools and methods:Dijkstra Algorithm": 26, - "CSETv1_Annotator-1:AI tools and methods:shortest-path algorithm": 26, - "CSETv1_Annotator-1:Data Inputs:GPS input": 26, - "CSETv1_Annotator-1:Data Inputs:traffic patterns": 26, - "CSETv1_Annotator-1:Data Inputs:user traffic reports": 26, - "CSETv1_Annotator-1:Location Country (two letters):GB": 26, - "CSETv1_Annotator-1:User Test in Controlled Conditions:yes": 26, - "CSETv1_Annotator-2:AI Task:Translation": 26, - "CSETv1_Annotator-2:Location Country (two letters):PS": 26, - "CSETv1_Annotator-3:AI Task:Translation": 26, - "CSETv1_Annotator-3:Location Country (two letters):PS": 26, - "GMF:Known AI Goal:Translation": 26, - "GMF:Potential AI Technical Failure:Misconfigured Aggregation": 26, - "GMF:Potential AI Technology:Image Classification": 26, - "GMF:Potential AI Technology:Intermediate modeling": 26, - "GMF:Potential AI Technology:Multimodal Learning": 26, - "CSETv0:Harm Type:Other": 25, - "CSETv0:Location:Ningbo, China": 25, - "CSETv0:Location:United States": 25, - "CSETv0:Named Entities:Dong Mingzhu": 25, - "CSETv0:Named Entities:Google Photos": 25, - "CSETv0:Named Entities:Gree Electric Appliances": 25, - "CSETv0:Named Entities:Ningbo": 25, - "CSETv0:Physical System:Other:CCTV cameras, displays": 25, - "CSETv0:System Developer:Ningbo traffic police": 25, - "CSETv0:Technology Purveyor:Ningbo traffic police": 25, - "CSETv1:AI Task:image classification": 25, - "CSETv1:AI Task:jaywalking detection": 25, - "CSETv1:Data Inputs:sex": 25, - "CSETv1:Infrastructure Sectors:information technology": 25, - "CSETv1:Lives Lost:2": 25, - "CSETv1:Operating Conditions:night": 25, - "CSETv1:Physical System Type:Tesla vehicle (Model 3, Model S, Model X)": 25, - "CSETv1:User Test in Controlled Conditions:yes": 25, - "CSETv1_Annotator-1:AI Task:jaywalking detection": 25, - "CSETv1_Annotator-1:Data Inputs:camera footage": 25, - "CSETv1_Annotator-1:Data Inputs:sex": 25, - "CSETv1_Annotator-1:Date of Incident Year:2009": 25, - "CSETv1_Annotator-1:Harm Distribution Basis:age": 25, - "CSETv1_Annotator-1:Physical System Type:Tesla vehicle (Model 3, Model S, Model X)": 25, - "CSETv1_Annotator-1:Physical System Type:autonomous car": 25, - "CSETv1_Annotator-2:AI Task:Detect Jaywalkers": 25, - "CSETv1_Annotator-2:Physical System Type:Camera": 25, - "CSETv1_Annotator-3:AI Harm Level:AI tangible harm issue": 25, - "GMF:Potential AI Technical Failure:Concept Drift": 25 - }, - "CSETv1.AI System": { - "yes": 1221, - "no": 269, - "maybe": 92 - }, - "CSETv1.Lives Lost": { - "0": 1476, - "1": 76, - "2": 25, - "18": 2, - "51": 1, - "130": 3, - "144": 12, - "189": 19 - }, - "GMF.Known AI Goal": { - "Autonomous Driving": 45, - "Face Recognition": 36, - "AI Voice Assistant": 35, - "Autonomous Drones": 31, - "Automatic Stock Trading": 30, - "Deepfake Video Generation": 29, - "Hate Speech Detection": 29, - "Threat Detection": 27, - "Translation": 26, - "Image Tagging": 24, - "Content Recommendation": 14, - "Content Search": 14, - "Market Forecasting": 14, - "NSFW Content Detection": 14, - "Chatbot": 13, - "Audio Localization": 11, - "Gunshot Detection": 11, - "Automatic Skill Assessment": 10, - "Scheduling": 10, - "Robotic Manipulation": 8, - "Visual Art Generation": 7, - "Smart Devices": 6, - "Substance Detection": 6, - "Code Generation": 5, - "Image Cropping": 5, - "Activity Tracking": 3, - "Question Answering": 3, - "Copyrighted Content Detection": 2, - "Data Grouping": 2, - "Social Media Content Generation": 2, - "Automated Content Curation": 1, - "Behavioral Modeling": 1, - "Financial Processing": 1, - "License Plate Recognition": 1, - "Recidivism Prediction": 1 - }, - "CSETv1.Data Inputs": { - "text": 443, - "images": 286, - "video": 234, - "sensor data": 147, - "audio": 143, - "lidar": 143, - "radar": 131, - "facial images": 100, - "camera input": 95, - "search queries": 68, - "speech": 64, - "video input": 61, - "voice": 61, - "photos": 58, - "road data": 57, - "spatial data": 52, - "infrared camera": 51, - "personal data": 51, - "image alt tags": 47, - "dot projector": 45, - "GPS": 44, - "income data": 39, - "tabular": 39, - "tax reports": 39, - "Amazon account information": 37, - "Facebook posts": 36, - "air quality": 36, - "Crime Records": 35, - "Resume": 33, - "traffic": 33, - "age": 31, - "personal information": 31, - "radar input": 30, - "stock data": 30, - "stock orders": 30, - "stock prices": 30, - "thermal imaging": 30, - "words": 30, - "Train position/status": 29, - "geospatial data": 29, - "track sensors": 29, - "Captions": 28, - "image captions": 28, - "laser input": 28, - "video captions": 28, - "satellite data": 27, - "sonar": 27, - "vibration detectors": 27, - "Youtube videos": 26, - "traffic patterns": 26, - "sex": 25, - "DAO token IDs": 24, - "GNSS Antennae input": 24, - "GPS input": 24, - "code": 24, - "inertial measurement unit input": 24, - "odometer input": 24, - "sales data": 24, - "traffic light signals": 24, - "user prompts": 24, - "user traffic reports": 24, - "previous criminal history": 23, - "user queries": 23, - "137 questionnaire responses": 22, - "Police reports": 22, - "camera footage": 22, - "criminal degree": 22, - "route specifications": 22, - "employee profile": 21, - "infrared images": 21, - "employee status": 20, - "sensor": 17, - "messages": 16, - "criminal record": 15, - "questionnaire responses": 15, - "keywords": 14, - "store camera footage": 14, - "surveillance footage": 14, - "census data": 13, - "maps": 13, - "names": 13, - "news articles": 13, - "past location of crime": 13, - "past time of crime": 13, - "past type of crime": 13, - "patrol shifts": 13, - "population density": 13, - "game data": 11, - "microphone inputs": 11, - "patient data": 11, - "schedules": 11, - "security footage": 11, - "selfies": 11, - "store traffic": 10, - "test scores": 10, - "worker profiles": 10, - "audio inputs": 9, - "search terms": 9, - "sensor input": 8, - "treatment cost": 7, - "Facebook accounts": 6, - "Youtube content": 6, - "credit report": 6, - "income": 6, - "motion": 6, - "personal debt": 6, - "personal finances": 6, - "temperature": 6, - "CDPH guidelines": 5, - "Facebook ads": 5, - "Instagram accounts": 5, - "Uber accounts": 5, - "department": 5, - "driver behavior data": 5, - "driver rating data": 5, - "job role": 5, - "medical personnel data": 5, - "148 million tax returns": 4, - "780000 audits": 4, - "Uber driver account information": 4, - "applicant data": 4, - "bluetooth": 4, - "cameras": 4, - "census demographics": 4, - "electronic health records": 4, - "employee activity": 4, - "employee performance data": 4, - "geographical information": 4, - "home valuations": 4, - "homeowner-submitted details": 4, - "house pictures": 4, - "images from social media": 4, - "images from websites": 4, - "property records": 4, - "real estate data": 4, - "scraped images": 4, - "social media accounts": 4, - "student data": 4, - "tax records": 4, - "taxpayer names": 4, - "taxpayer profile": 4, - "user data": 4, - "Employee Info": 3, - "GPA": 3, - "Product Info": 3, - "RFID": 3, - "Time off Task (TOT)": 3, - "Unclear": 3, - "Warehouse data": 3, - "creatinine levels": 3, - "email addresses": 3, - "ethical judgments": 3, - "ethical scenarios": 3, - "medical data": 3, - "race": 3, - "student grades": 3, - "usernames": 3, - "voice input": 3, - "water samples": 3, - "worker performance data": 3, - "Amazon products": 2, - "Drake songs": 2, - "Drake voice": 2, - "The Weeknd songs": 2, - "The Weeknd voice": 2, - "TikTok posts": 2, - "area of research interest": 2, - "biometrics": 2, - "body scans": 2, - "faculty advisor": 2, - "family relationships": 2, - "letter of recommendation": 2, - "letters of recommendation": 2, - "license plate images": 2, - "patient information": 2, - "photos in Google photos": 2, - "purchase history": 2, - "reliability index": 2, - "shift data": 2, - "songs": 2, - "statement of interest": 2, - "university name": 2, - "university previously attended": 2, - "websites": 2, - "286 questions": 1, - "Amazon items": 1, - "Amazon purchases": 1, - "Burmese text": 1, - "Facebook comments": 1, - "Facebook likes": 1, - "Facebook reactions": 1, - "Facebook reshares": 1, - "Instagram posts": 1, - "Internet usage": 1, - "SAT and ACT scores": 1, - "TikTok accounts": 1, - "Twitter posts": 1, - "acceleration": 1, - "advertisements": 1, - "alleged crimes": 1, - "arrest photos": 1, - "article text": 1, - "audio recordings": 1, - "biometric data": 1, - "breaking": 1, - "credits attempted": 1, - "credits completed": 1, - "dates of significance": 1, - "delivery times": 1, - "diagnosis": 1, - "driver performance metrics": 1, - "emails": 1, - "estimated skills": 1, - "health data": 1, - "health measures": 1, - "high school percentile": 1, - "in-van camera input": 1, - "income statements": 1, - "ingredients": 1, - "job application": 1, - "living situation": 1, - "marketing material": 1, - "national IDs": 1, - "phone calls": 1, - "photo IDs": 1, - "physical function": 1, - "predicted final grades": 1, - "property data": 1, - "school attended": 1, - "school principal assessments": 1, - "schools' historical IB results": 1, - "seatbelt usage": 1, - "social media posts": 1, - "speed": 1, - "standardized test results": 1, - "statewide average test scores": 1, - "student evaluations": 1, - "student test results": 1, - "symptoms": 1, - "texting": 1, - "uploaded images": 1, - "user comments": 1, - "user engagement": 1, - "user reports": 1, - "voice recordings": 1, - "webpage content": 1, - "website code": 1, - "welfare recipient data": 1, - "x-ray images": 1, - "youtube history": 1, - "zip code": 1 - }, - "is_incident_report": { - "true": 3871 - }, - "epoch_incident_date": { - "433382400": 27, - "706752000": 2, - "828489600": 1, - "889056000": 4, - "921542400": 3, - "1048291200": 1, - "1049241600": 4, - "1054425600": 3, - "1057017600": 2, - "1140825600": 3, - "1185408000": 7, - "1211500800": 24, - "1214870400": 1, - "1216339200": 4, - "1238371200": 1, - "1251763200": 1, - "1273276800": 30, - "1301961600": 29, - "1303084800": 1, - "1316476800": 2, - "1325376000": 1, - "1325721600": 1, - "1330128000": 7, - "1336089600": 4, - "1343779200": 10, - "1349740800": 11, - "1354320000": 2, - "1355097600": 2, - "1358899200": 27, - "1375142400": 2, - "1377993600": 5, - "1379030400": 3, - "1380585600": 12, - "1385510400": 1, - "1388534400": 1, - "1390262400": 6, - "1393891200": 5, - "1397606400": 2, - "1399507200": 2, - "1405382400": 27, - "1407974400": 10, - "1408060800": 10, - "1409011200": 2, - "1412121600": 2, - "1413590400": 20, - "1418083200": 3, - "1419206400": 1, - "1425168000": 16, - "1427155200": 1, - "1427846400": 3, - "1428105600": 11, - "1430438400": 2, - "1431302400": 11, - "1431993600": 14, - "1433116800": 2, - "1433289600": 24, - "1434240000": 1, - "1435708800": 39, - "1435795200": 12, - "1436745600": 12, - "1436918400": 1, - "1438387200": 1, - "1441065600": 13, - "1443139200": 5, - "1446508800": 22, - "1447372800": 3, - "1447804800": 13, - "1449273600": 35, - "1451606400": 1, - "1453248000": 4, - "1454284800": 1, - "1454457600": 1, - "1455062400": 7, - "1456790400": 8, - "1458777600": 28, - "1459382400": 18, - "1459468800": 4, - "1459814400": 4, - "1460160000": 1, - "1460678400": 1, - "1463961600": 37, - "1464220800": 3, - "1464739200": 10, - "1464825600": 12, - "1465948800": 5, - "1466121600": 24, - "1467244800": 22, - "1467331200": 29, - "1468281600": 27, - "1469059200": 1, - "1470787200": 33, - "1472688000": 5, - "1473033600": 10, - "1473120000": 9, - "1473379200": 13, - "1474848000": 28, - "1475280000": 2, - "1475884800": 29, - "1479254400": 2, - "1481068800": 22, - "1481760000": 1, - "1482364800": 1, - "1483056000": 16, - "1483142400": 4, - "1483228800": 1, - "1484438400": 4, - "1485907200": 1, - "1487894400": 6, - "1488153600": 9, - "1488326400": 2, - "1489363200": 3, - "1489536000": 2, - "1491436800": 5, - "1491523200": 2, - "1491782400": 3, - "1492041600": 10, - "1492214400": 1, - "1493078400": 23, - "1493337600": 1, - "1493596800": 1, - "1494201600": 1, - "1495152000": 3, - "1496448000": 8, - "1497484800": 10, - "1498867200": 35, - "1499040000": 2, - "1499644800": 7, - "1500249600": 30, - "1501545600": 2, - "1501632000": 16, - "1502755200": 2, - "1503705600": 4, - "1504742400": 1, - "1505260800": 45, - "1505433600": 1, - "1505692800": 1, - "1505952000": 2, - "1506470400": 1, - "1507593600": 5, - "1507766400": 5, - "1508198400": 26, - "1508976000": 7, - "1509494400": 1, - "1510099200": 24, - "1510185600": 4, - "1510704000": 8, - "1511222400": 1, - "1512259200": 29, - "1512518400": 22, - "1515715200": 3, - "1515974400": 2, - "1516579200": 5, - "1516838400": 1, - "1517443200": 1, - "1518652800": 2, - "1521331200": 25, - "1521763200": 14, - "1522627200": 28, - "1523404800": 4, - "1523491200": 3, - "1524182400": 2, - "1524441600": 2, - "1524960000": 4, - "1525132800": 6, - "1525910400": 4, - "1525996800": 1, - "1527292800": 4, - "1527552000": 4, - "1527811200": 1, - "1528675200": 3, - "1530489600": 1, - "1530662400": 1, - "1530921600": 4, - "1532044800": 2, - "1532131200": 3, - "1532563200": 1, - "1533081600": 3, - "1534118400": 4, - "1534291200": 5, - "1535760000": 6, - "1538352000": 2, - "1540598400": 19, - "1541462400": 25, - "1541635200": 5, - "1541721600": 2, - "1542931200": 2, - "1543363200": 2, - "1543622400": 25, - "1543795200": 1, - "1543968000": 17, - "1546300800": 1, - "1548806400": 4, - "1548979200": 3, - "1549238400": 3, - "1550102400": 3, - "1551398400": 6, - "1552262400": 3, - "1552608000": 1, - "1554076800": 7, - "1554249600": 3, - "1559260800": 3, - "1559347200": 5, - "1559520000": 2, - "1561161600": 8, - "1561420800": 2, - "1561939200": 3, - "1562112000": 3, - "1562371200": 3, - "1562716800": 1, - "1562803200": 2, - "1564531200": 3, - "1566604800": 1, - "1567728000": 2, - "1569715200": 3, - "1570147200": 5, - "1570492800": 3, - "1570579200": 1, - "1571097600": 6, - "1571875200": 7, - "1572998400": 4, - "1573430400": 6, - "1573516800": 6, - "1573776000": 4, - "1574985600": 4, - "1575676800": 2, - "1577577600": 8, - "1577836800": 6, - "1577923200": 1, - "1579046400": 4, - "1579305600": 1, - "1579564800": 6, - "1580342400": 11, - "1580947200": 1, - "1581033600": 2, - "1582070400": 3, - "1582156800": 3, - "1582502400": 4, - "1583712000": 4, - "1584316800": 4, - "1584403200": 1, - "1584921600": 2, - "1585267200": 4, - "1585699200": 1, - "1586822400": 2, - "1586908800": 1, - "1588032000": 3, - "1588377600": 2, - "1590883200": 9, - "1590969600": 4, - "1591401600": 12, - "1591574400": 2, - "1591747200": 1, - "1591833600": 2, - "1592006400": 2, - "1592179200": 2, - "1592352000": 1, - "1592611200": 8, - "1592956800": 4, - "1593216000": 1, - "1593302400": 6, - "1593561600": 9, - "1593993600": 1, - "1595030400": 1, - "1595894400": 8, - "1596412800": 1, - "1596585600": 2, - "1596672000": 3, - "1596758400": 3, - "1597276800": 8, - "1598918400": 1, - "1600128000": 2, - "1600387200": 5, - "1600560000": 1, - "1601337600": 3, - "1601683200": 3, - "1601942400": 2, - "1602028800": 1, - "1602115200": 2, - "1602201600": 2, - "1602720000": 1, - "1603152000": 5, - "1603238400": 2, - "1603324800": 1, - "1603497600": 2, - "1603929600": 1, - "1604188800": 3, - "1605052800": 4, - "1605398400": 1, - "1606435200": 2, - "1607040000": 2, - "1607212800": 1, - "1607990400": 2, - "1608249600": 5, - "1608681600": 13, - "1608768000": 1, - "1608854400": 2, - "1609027200": 1, - "1609459200": 4, - "1610236800": 4, - "1610323200": 1, - "1610496000": 2, - "1610668800": 1, - "1611187200": 2, - "1612137600": 1, - "1612483200": 3, - "1613001600": 1, - "1613088000": 1, - "1613433600": 1, - "1614124800": 1, - "1614297600": 2, - "1614556800": 1, - "1614643200": 4, - "1615766400": 2, - "1615939200": 7, - "1617235200": 3, - "1618272000": 1, - "1618617600": 9, - "1618963200": 1, - "1619568000": 1, - "1619654400": 2, - "1619827200": 8, - "1620259200": 3, - "1621296000": 7, - "1621728000": 1, - "1622160000": 6, - "1622592000": 7, - "1622678400": 4, - "1623456000": 2, - "1624924800": 5, - "1625097600": 1, - "1625184000": 1, - "1625356800": 4, - "1625875200": 3, - "1625961600": 2, - "1626134400": 2, - "1626393600": 4, - "1626739200": 1, - "1626998400": 3, - "1627603200": 1, - "1627776000": 14, - "1627948800": 4, - "1628985600": 4, - "1630454400": 3, - "1631836800": 3, - "1632096000": 2, - "1633046400": 4, - "1633478400": 1, - "1634256000": 3, - "1634515200": 1, - "1634774400": 1, - "1634860800": 3, - "1635379200": 10, - "1635724800": 8, - "1635811200": 4, - "1635897600": 1, - "1636243200": 1, - "1636502400": 5, - "1637452800": 3, - "1638748800": 18, - "1638921600": 1, - "1639180800": 7, - "1640390400": 2, - "1640476800": 2, - "1640563200": 2, - "1640995200": 1, - "1642118400": 1, - "1642204800": 10, - "1642809600": 3, - "1643155200": 1, - "1643932800": 5, - "1644451200": 3, - "1644537600": 6, - "1644883200": 1, - "1645574400": 1, - "1645747200": 4, - "1646006400": 4, - "1646092800": 7, - "1646179200": 1, - "1646265600": 1, - "1646611200": 2, - "1647302400": 3, - "1647388800": 9, - "1647475200": 2, - "1647907200": 2, - "1648771200": 8, - "1648944000": 1, - "1649116800": 2, - "1649203200": 2, - "1649808000": 1, - "1650326400": 5, - "1650412800": 3, - "1650499200": 8, - "1651968000": 2, - "1652400000": 2, - "1652659200": 17, - "1652832000": 4, - "1653436800": 3, - "1654041600": 1, - "1654214400": 12, - "1654905600": 4, - "1654992000": 2, - "1656374400": 1, - "1658102400": 1, - "1658361600": 5, - "1658448000": 3, - "1658620800": 3, - "1658793600": 2, - "1659830400": 3, - "1660521600": 3, - "1660694400": 1, - "1661040000": 2, - "1661126400": 3, - "1661385600": 2, - "1661731200": 1, - "1662508800": 2, - "1662854400": 2, - "1663027200": 3, - "1663200000": 18, - "1663891200": 1, - "1664755200": 2, - "1664841600": 8, - "1665446400": 1, - "1665878400": 5, - "1667952000": 1, - "1668124800": 14, - "1668470400": 6, - "1668902400": 9, - "1669075200": 6, - "1669248000": 8, - "1669334400": 9, - "1669507200": 1, - "1669766400": 18, - "1669852800": 15, - "1671408000": 23, - "1671580800": 26, - "1672185600": 10, - "1672531200": 5, - "1672704000": 8, - "1672876800": 1, - "1673395200": 9, - "1674000000": 4, - "1674172800": 3, - "1674259200": 2, - "1675036800": 19, - "1675296000": 2, - "1675382400": 6, - "1675641600": 5, - "1675728000": 18, - "1675814400": 4, - "1675987200": 2, - "1676160000": 11, - "1676332800": 12, - "1676419200": 3, - "1676505600": 23, - "1676851200": 4, - "1677024000": 1, - "1677456000": 2, - "1677542400": 22, - "1677801600": 2, - "1677974400": 5, - "1678492800": 2, - "1678838400": 5, - "1679011200": 2, - "1679270400": 2, - "1679356800": 11, - "1679529600": 2, - "1679616000": 5, - "1679875200": 7, - "1680048000": 2, - "1680220800": 5, - "1680307200": 3, - "1680912000": 2, - "1681430400": 7, - "1681689600": 2, - "1682380800": 1, - "1682899200": 1, - "1683072000": 2, - "1683158400": 58, - "1683763200": 22, - "1683849600": 1, - "1684108800": 10, - "1684368000": 1, - "1684627200": 1, - "1684713600": 19, - "1684886400": 2, - "1685232000": 1, - "1685318400": 46, - "1685577600": 3, - "1685923200": 2, - "1686096000": 1, - "1686528000": 1, - "1686614400": 4, - "1687392000": 2, - "1687478400": 2, - "1687564800": 1, - "1687737600": 1, - "1687910400": 1, - "1688083200": 3, - "1688342400": 1, - "1688515200": 1, - "1689033600": 2, - "1689552000": 1, - "1689897600": 1, - "1690156800": 3, - "1691452800": 4, - "1691625600": 1, - "1691712000": 1, - "1691971200": 2, - "1692144000": 2, - "1693094400": 1, - "1693353600": 1, - "1693612800": 1, - "1694390400": 1, - "1694908800": 10, - "1695081600": 1, - "1696204800": 27, - "1696377600": 1, - "1696636800": 14, - "1696723200": 10, - "1697414400": 44, - "1697500800": 1, - "1698105600": 1, - "1698278400": 3, - "1698710400": 17, - "1698883200": 1, - "1699401600": 2, - "1699488000": 3, - "1699574400": 3, - "1700697600": 1, - "1701043200": 43, - "1701388800": 3, - "1701820800": 3, - "1702252800": 8, - "1702339200": 12, - "1702512000": 3, - "1702684800": 1, - "1702857600": 6, - "1703030400": 32, - "1703548800": 30, - "1703635200": 1, - "1703808000": 1, - "1704758400": 8, - "1704931200": 1, - "1705017600": 5, - "1705104000": 1, - "1705276800": 4, - "1705536000": 2, - "1705795200": 14, - "1705968000": 2, - "1706054400": 31, - "1706400000": 17, - "1706572800": 2, - "1706745600": 3, - "1706832000": 21, - "1707091200": 1, - "1707264000": 2, - "1707350400": 1, - "1707609600": 1, - "1707696000": 1, - "1707782400": 6, - "1707868800": 7, - "1708128000": 1, - "1708214400": 6, - "1708300800": 1, - "1708387200": 17, - "1708473600": 35, - "1708560000": 1, - "1709078400": 34, - "1709251200": 6, - "1709510400": 1, - "1709683200": 3, - "1710374400": 7, - "1710979200": 1, - "1711065600": 3, - "1711152000": 3, - "1711411200": 9, - "1711584000": 2, - "1711670400": 2, - "1711929600": 3, - "1712016000": 3, - "1712102400": 9, - "1712188800": 2, - "1712275200": 1, - "1712361600": 3, - "1712448000": 2, - "1712534400": 1, - "1713052800": 1, - "1713139200": 1, - "1713225600": 2, - "1713484800": 10, - "1713830400": 1, - "1713916800": 3, - "1714003200": 2, - "1714089600": 4, - "1714348800": 2, - "1714694400": 3, - "1715558400": 2, - "1715644800": 9, - "1715731200": 2, - "1715817600": 1, - "1716163200": 16, - "1716249600": 1, - "1716422400": 4, - "1716508800": 16, - "1716595200": 2, - "1716681600": 1, - "1716940800": 1, - "1717027200": 2, - "1717113600": 2, - "1717200000": 3, - "1717459200": 1, - "1717632000": 1, - "1717718400": 5, - "1717804800": 2, - "1717891200": 2, - "1718150400": 2, - "1718236800": 1, - "1718668800": 3, - "1718755200": 1, - "1718841600": 1, - "1718928000": 2, - "1719014400": 1, - "1719100800": 5, - "1719273600": 1, - "1719360000": 4, - "1719446400": 2, - "1719619200": 1, - "1719792000": 4, - "1719878400": 2, - "1719964800": 2, - "1720051200": 2, - "1720310400": 2, - "1720569600": 1, - "1720828800": 1, - "1721088000": 1, - "1721260800": 1, - "1721520000": 2, - "1721606400": 1, - "1721692800": 1, - "1721865600": 2, - "1721952000": 3, - "1722211200": 1, - "1722297600": 5, - "1722902400": 1, - "1723075200": 1, - "1723507200": 3, - "1723593600": 4, - "1723766400": 3, - "1723939200": 4, - "1724112000": 1, - "1724198400": 1, - "1724371200": 1, - "1724803200": 7, - "1725235200": 2, - "1725321600": 2, - "1725408000": 3, - "1725753600": 1, - "1725840000": 2, - "1726185600": 2, - "1726617600": 1, - "1726704000": 12, - "1727136000": 8, - "1727222400": 7, - "1727481600": 2, - "1727740800": 3, - "1727827200": 13, - "1728172800": 1, - "1728259200": 20, - "1728345600": 2, - "1728432000": 1, - "1728604800": 1, - "1728777600": 3, - "1728950400": 7, - "1729036800": 11, - "1729209600": 2, - "1729468800": 2, - "1729641600": 1, - "1729728000": 1, - "1729900800": 1, - "1730505600": 1, - "1731456000": 1, - "1732060800": 2 - }, - "CSETv1.AI Harm Level": { - "none": 1003, - "AI tangible harm event": 373, - "AI tangible harm near-miss": 104, - "AI tangible harm issue": 66, - "unclear": 36 - }, - "CSETv1.Tangible Harm": { - "tangible harm definitively occurred": 678, - "no tangible harm, near-miss, or issue": 618, - "imminent risk of tangible harm (near miss) did occur": 180, - "non-imminent risk of tangible harm (an issue) occurred": 97, - "unclear": 9 - }, - "epoch_date_published": { - "1062028800": 1, - "1084579200": 1, - "1085011200": 1, - "1095292800": 1, - "1102723200": 1, - "1170806400": 1, - "1211500800": 1, - "1214870400": 1, - "1219104000": 1, - "1238803200": 1, - "1239494400": 5, - "1239580800": 7, - "1239667200": 5, - "1239840000": 1, - "1240012800": 1, - "1240531200": 1, - "1242259200": 1, - "1273276800": 1, - "1282176000": 1, - "1285804800": 1, - "1301961600": 2, - "1302048000": 1, - "1303430400": 1, - "1303862400": 1, - "1304553600": 1, - "1308009600": 1, - "1316476800": 1, - "1325721600": 2, - "1329350400": 2, - "1330128000": 1, - "1330300800": 1, - "1330387200": 1, - "1330560000": 1, - "1332720000": 1, - "1332892800": 1, - "1347753600": 1, - "1348099200": 1, - "1348704000": 1, - "1351641600": 1, - "1352160000": 1, - "1357948800": 1, - "1359417600": 1, - "1359676800": 1, - "1359936000": 5, - "1360022400": 5, - "1360108800": 4, - "1360195200": 2, - "1360800000": 1, - "1364860800": 1, - "1366070400": 1, - "1366243200": 1, - "1368403200": 1, - "1368489600": 1, - "1369785600": 1, - "1375142400": 1, - "1375747200": 1, - "1376697600": 1, - "1380153600": 1, - "1382140800": 1, - "1382572800": 1, - "1390262400": 1, - "1396569600": 1, - "1396656000": 1, - "1398211200": 1, - "1398643200": 1, - "1399334400": 1, - "1400630400": 1, - "1400803200": 2, - "1405382400": 1, - "1407283200": 2, - "1407369600": 1, - "1407974400": 1, - "1411689600": 1, - "1413590400": 1, - "1415318400": 1, - "1418083200": 1, - "1419379200": 1, - "1419811200": 2, - "1423267200": 1, - "1427932800": 1, - "1428105600": 1, - "1428537600": 2, - "1428624000": 2, - "1428710400": 1, - "1429056000": 1, - "1429142400": 1, - "1429574400": 2, - "1429660800": 7, - "1429747200": 5, - "1430092800": 1, - "1430784000": 1, - "1430870400": 1, - "1431302400": 1, - "1431993600": 1, - "1433289600": 1, - "1433376000": 1, - "1435276800": 3, - "1435363200": 1, - "1435449600": 1, - "1435536000": 1, - "1435622400": 1, - "1435708800": 9, - "1435795200": 28, - "1435881600": 1, - "1435968000": 1, - "1436140800": 2, - "1436313600": 1, - "1436572800": 1, - "1436745600": 2, - "1437350400": 1, - "1437436800": 6, - "1437523200": 2, - "1437868800": 1, - "1438560000": 1, - "1439078400": 1, - "1439424000": 5, - "1439510400": 2, - "1439683200": 1, - "1440633600": 1, - "1441065600": 1, - "1443052800": 2, - "1443571200": 1, - "1446422400": 1, - "1446508800": 6, - "1446595200": 1, - "1446681600": 1, - "1446854400": 1, - "1447286400": 1, - "1447804800": 1, - "1448409600": 1, - "1449273600": 1, - "1451606400": 1, - "1452729600": 1, - "1454112000": 1, - "1454457600": 1, - "1454544000": 1, - "1455062400": 2, - "1455235200": 1, - "1455840000": 1, - "1456444800": 1, - "1456704000": 3, - "1456790400": 9, - "1457049600": 1, - "1457568000": 5, - "1458000000": 1, - "1458691200": 1, - "1458777600": 6, - "1458864000": 7, - "1458950400": 3, - "1459036800": 2, - "1459296000": 1, - "1459382400": 1, - "1459555200": 1, - "1459987200": 2, - "1460073600": 2, - "1460419200": 1, - "1461283200": 1, - "1462060800": 1, - "1463961600": 6, - "1464048000": 1, - "1464220800": 3, - "1464393600": 1, - "1464825600": 2, - "1464912000": 3, - "1464998400": 2, - "1465084800": 1, - "1465171200": 1, - "1465257600": 2, - "1465344000": 2, - "1465430400": 4, - "1465516800": 7, - "1466121600": 1, - "1466208000": 3, - "1466294400": 1, - "1466380800": 1, - "1466553600": 1, - "1466812800": 2, - "1466899200": 1, - "1467244800": 1, - "1467331200": 13, - "1467417600": 7, - "1467504000": 1, - "1467590400": 1, - "1467676800": 1, - "1467849600": 1, - "1467936000": 1, - "1468195200": 1, - "1468281600": 4, - "1468368000": 12, - "1468454400": 6, - "1468540800": 3, - "1468800000": 1, - "1469059200": 1, - "1469232000": 1, - "1469491200": 1, - "1470441600": 2, - "1470700800": 1, - "1470787200": 2, - "1470873600": 1, - "1471392000": 1, - "1471564800": 1, - "1471824000": 2, - "1471996800": 1, - "1472083200": 1, - "1472428800": 1, - "1472601600": 1, - "1473033600": 1, - "1473120000": 2, - "1473206400": 1, - "1473292800": 4, - "1473379200": 5, - "1473465600": 1, - "1473811200": 2, - "1474243200": 1, - "1474329600": 1, - "1474675200": 1, - "1474761600": 2, - "1474848000": 5, - "1475798400": 1, - "1475884800": 1, - "1476057600": 1, - "1476144000": 1, - "1477267200": 1, - "1477872000": 1, - "1479427200": 4, - "1480550400": 1, - "1480636800": 1, - "1480723200": 1, - "1480982400": 1, - "1481068800": 10, - "1481155200": 9, - "1481760000": 1, - "1481846400": 2, - "1482019200": 1, - "1482278400": 1, - "1482364800": 1, - "1482537600": 1, - "1482883200": 3, - "1483056000": 1, - "1483142400": 4, - "1483315200": 1, - "1483401600": 2, - "1483488000": 5, - "1483574400": 3, - "1483660800": 7, - "1483747200": 2, - "1483833600": 2, - "1483920000": 10, - "1484006400": 4, - "1484092800": 3, - "1484524800": 3, - "1484611200": 2, - "1485216000": 1, - "1485561600": 1, - "1486339200": 1, - "1486425600": 2, - "1486857600": 1, - "1487894400": 2, - "1487980800": 2, - "1488067200": 1, - "1488153600": 3, - "1488240000": 1, - "1488326400": 5, - "1488499200": 2, - "1488585600": 1, - "1488931200": 2, - "1489622400": 1, - "1489795200": 1, - "1490227200": 3, - "1490572800": 1, - "1491264000": 1, - "1491782400": 4, - "1491868800": 1, - "1492041600": 1, - "1492128000": 1, - "1492473600": 1, - "1492560000": 1, - "1493078400": 6, - "1493164800": 1, - "1493251200": 1, - "1493424000": 1, - "1493596800": 1, - "1493769600": 1, - "1493856000": 1, - "1494201600": 1, - "1495152000": 2, - "1495238400": 2, - "1496102400": 1, - "1497225600": 1, - "1497312000": 2, - "1497830400": 1, - "1498003200": 1, - "1498089600": 2, - "1498694400": 1, - "1499040000": 1, - "1499299200": 1, - "1499644800": 3, - "1499731200": 5, - "1499817600": 2, - "1499904000": 2, - "1499990400": 3, - "1500249600": 2, - "1500336000": 21, - "1500422400": 10, - "1500595200": 2, - "1500854400": 1, - "1501027200": 1, - "1501200000": 1, - "1501545600": 1, - "1501718400": 7, - "1501804800": 5, - "1501891200": 3, - "1502064000": 1, - "1502150400": 1, - "1502236800": 2, - "1502323200": 11, - "1502496000": 1, - "1502755200": 1, - "1502928000": 1, - "1503014400": 2, - "1503187200": 1, - "1503273600": 2, - "1503360000": 1, - "1503446400": 1, - "1504224000": 1, - "1504742400": 2, - "1504915200": 1, - "1505088000": 1, - "1505260800": 4, - "1505347200": 4, - "1505433600": 1, - "1505692800": 6, - "1505779200": 3, - "1505952000": 2, - "1506038400": 2, - "1506470400": 1, - "1506556800": 1, - "1506729600": 1, - "1507248000": 1, - "1507507200": 1, - "1507680000": 3, - "1507766400": 3, - "1507852800": 2, - "1507939200": 1, - "1508198400": 1, - "1508630400": 2, - "1508716800": 4, - "1508803200": 12, - "1508889600": 4, - "1508976000": 6, - "1509062400": 3, - "1509235200": 2, - "1509321600": 1, - "1509408000": 5, - "1509494400": 3, - "1509667200": 1, - "1509840000": 1, - "1509926400": 2, - "1510012800": 1, - "1510099200": 2, - "1510185600": 14, - "1510272000": 7, - "1510358400": 2, - "1510531200": 8, - "1510617600": 6, - "1510704000": 4, - "1510876800": 1, - "1510963200": 1, - "1511308800": 1, - "1511740800": 1, - "1511827200": 3, - "1512000000": 1, - "1512086400": 1, - "1512259200": 1, - "1512518400": 2, - "1512604800": 4, - "1512691200": 11, - "1512777600": 4, - "1513036800": 2, - "1513123200": 4, - "1513209600": 1, - "1513296000": 2, - "1513382400": 1, - "1513468800": 1, - "1513555200": 1, - "1513641600": 5, - "1513728000": 11, - "1513814400": 7, - "1514764800": 2, - "1515628800": 3, - "1515715200": 2, - "1516060800": 2, - "1516147200": 7, - "1516233600": 6, - "1516579200": 3, - "1516752000": 4, - "1516838400": 2, - "1517529600": 2, - "1517788800": 3, - "1518566400": 3, - "1519776000": 2, - "1521417600": 2, - "1521504000": 7, - "1521763200": 3, - "1522281600": 6, - "1522454400": 6, - "1522627200": 5, - "1522800000": 2, - "1523404800": 5, - "1523491200": 3, - "1523577600": 4, - "1523664000": 3, - "1523836800": 4, - "1523923200": 3, - "1524009600": 5, - "1524096000": 2, - "1524182400": 2, - "1524528000": 2, - "1525132800": 3, - "1525478400": 2, - "1525651200": 4, - "1525737600": 3, - "1526342400": 2, - "1526428800": 2, - "1526515200": 3, - "1527120000": 2, - "1527206400": 2, - "1527465600": 2, - "1527552000": 4, - "1528070400": 3, - "1528156800": 2, - "1528243200": 2, - "1528329600": 11, - "1528416000": 5, - "1528761600": 2, - "1528848000": 5, - "1529020800": 2, - "1529539200": 5, - "1529625600": 8, - "1529712000": 3, - "1529884800": 6, - "1531180800": 2, - "1532390400": 2, - "1532476800": 2, - "1533686400": 2, - "1536192000": 2, - "1536624000": 2, - "1537228800": 2, - "1537401600": 3, - "1537488000": 3, - "1537833600": 2, - "1537920000": 4, - "1538092800": 2, - "1538438400": 2, - "1539129600": 7, - "1539216000": 14, - "1539302400": 4, - "1540857600": 2, - "1541635200": 5, - "1541808000": 2, - "1542326400": 2, - "1542844800": 2, - "1542931200": 8, - "1543017600": 2, - "1543190400": 2, - "1543276800": 6, - "1543363200": 6, - "1543449600": 4, - "1543622400": 3, - "1543708800": 3, - "1543795200": 6, - "1543881600": 3, - "1543968000": 5, - "1544054400": 11, - "1544140800": 6, - "1544486400": 2, - "1547510400": 2, - "1549324800": 2, - "1553126400": 2, - "1553731200": 2, - "1555113600": 4, - "1555200000": 7, - "1555977600": 3, - "1557360000": 2, - "1557964800": 2, - "1561420800": 2, - "1568592000": 2, - "1570147200": 3, - "1570665600": 2, - "1571270400": 3, - "1571875200": 3, - "1572998400": 4, - "1573516800": 4, - "1574121600": 2, - "1579305600": 3, - "1580256000": 2, - "1581292800": 2, - "1582070400": 2, - "1582588800": 2, - "1582675200": 3, - "1582761600": 3, - "1588291200": 2, - "1588896000": 2, - "1590105600": 2, - "1590710400": 3, - "1590796800": 4, - "1590969600": 5, - "1591056000": 2, - "1591660800": 3, - "1591747200": 2, - "1592956800": 3, - "1594339200": 2, - "1595203200": 2, - "1595980800": 2, - "1596067200": 3, - "1596153600": 2, - "1596499200": 5, - "1596585600": 3, - "1597104000": 2, - "1597190400": 2, - "1597622400": 3, - "1597795200": 2, - "1597881600": 2, - "1597968000": 2, - "1598486400": 2, - "1599091200": 2, - "1599609600": 2, - "1599696000": 2, - "1599782400": 2, - "1600646400": 2, - "1601337600": 2, - "1601942400": 3, - "1602028800": 2, - "1602201600": 4, - "1602720000": 2, - "1603152000": 2, - "1603324800": 2, - "1603670400": 3, - "1605744000": 2, - "1605830400": 2, - "1606262400": 2, - "1606521600": 3, - "1606780800": 2, - "1608249600": 4, - "1608422400": 2, - "1608768000": 2, - "1609027200": 3, - "1609286400": 2, - "1609545600": 2, - "1609718400": 2, - "1609804800": 2, - "1610236800": 2, - "1610323200": 2, - "1610409600": 3, - "1610496000": 4, - "1610582400": 3, - "1610928000": 2, - "1611014400": 2, - "1611878400": 4, - "1612137600": 2, - "1612310400": 2, - "1612569600": 2, - "1612828800": 2, - "1613001600": 2, - "1613433600": 2, - "1613692800": 2, - "1614211200": 2, - "1614556800": 2, - "1614643200": 2, - "1615334400": 2, - "1615507200": 3, - "1615939200": 3, - "1616025600": 3, - "1616112000": 2, - "1617321600": 2, - "1617926400": 3, - "1618704000": 3, - "1618790400": 2, - "1618876800": 2, - "1619395200": 2, - "1619568000": 3, - "1621382400": 2, - "1621468800": 2, - "1622592000": 3, - "1622678400": 3, - "1622764800": 3, - "1624838400": 2, - "1624924800": 4, - "1625184000": 2, - "1625443200": 3, - "1626220800": 3, - "1626652800": 4, - "1626825600": 2, - "1627257600": 2, - "1627603200": 2, - "1628208000": 2, - "1628467200": 2, - "1629331200": 2, - "1629849600": 2, - "1632096000": 2, - "1634083200": 4, - "1634601600": 2, - "1634688000": 3, - "1634860800": 3, - "1635811200": 2, - "1635897600": 3, - "1636329600": 4, - "1636416000": 3, - "1636502400": 2, - "1636588800": 3, - "1636761600": 2, - "1636934400": 6, - "1637193600": 2, - "1637452800": 3, - "1638144000": 3, - "1638403200": 2, - "1638662400": 2, - "1639008000": 4, - "1639094400": 2, - "1639353600": 2, - "1639440000": 8, - "1639526400": 4, - "1639612800": 4, - "1640649600": 2, - "1641427200": 2, - "1641772800": 3, - "1642464000": 3, - "1643241600": 3, - "1643760000": 4, - "1643932800": 2, - "1644192000": 2, - "1644278400": 2, - "1644796800": 2, - "1645574400": 2, - "1646006400": 3, - "1646092800": 2, - "1646611200": 3, - "1646697600": 5, - "1646784000": 3, - "1647388800": 5, - "1647475200": 7, - "1647820800": 3, - "1647907200": 3, - "1647993600": 2, - "1648166400": 3, - "1648425600": 2, - "1648512000": 3, - "1648684800": 2, - "1648771200": 2, - "1649030400": 2, - "1649116800": 2, - "1649548800": 4, - "1649635200": 2, - "1650585600": 3, - "1650672000": 3, - "1650844800": 4, - "1650931200": 2, - "1653177600": 2, - "1653609600": 2, - "1653868800": 2, - "1654128000": 2, - "1654214400": 2, - "1654473600": 4, - "1655164800": 2, - "1655251200": 2, - "1657065600": 2, - "1657152000": 3, - "1657238400": 3, - "1657497600": 2, - "1658102400": 2, - "1658361600": 2, - "1658448000": 5, - "1658620800": 4, - "1658707200": 3, - "1658793600": 3, - "1658880000": 2, - "1658966400": 3, - "1659657600": 2, - "1659744000": 3, - "1659916800": 2, - "1660089600": 2, - "1660521600": 2, - "1661126400": 4, - "1661212800": 2, - "1661299200": 2, - "1661385600": 2, - "1661472000": 2, - "1661904000": 2, - "1661990400": 2, - "1662076800": 2, - "1662940800": 2, - "1663027200": 2, - "1663113600": 2, - "1663286400": 4, - "1663632000": 2, - "1664755200": 2, - "1664841600": 3, - "1665014400": 3, - "1665100800": 2, - "1665360000": 2, - "1665619200": 3, - "1666224000": 2, - "1667174400": 2, - "1667260800": 2, - "1668124800": 2, - "1668643200": 2, - "1668729600": 2, - "1669075200": 2, - "1669852800": 6, - "1670025600": 2, - "1670198400": 5, - "1670284800": 4, - "1670371200": 3, - "1670457600": 2, - "1670544000": 3, - "1670803200": 2, - "1670976000": 5, - "1671148800": 3, - "1671408000": 4, - "1671494400": 4, - "1671580800": 9, - "1671667200": 5, - "1672272000": 2, - "1672444800": 2, - "1672617600": 6, - "1672704000": 4, - "1672876800": 3, - "1672963200": 5, - "1673049600": 2, - "1673136000": 2, - "1673222400": 9, - "1673308800": 5, - "1673395200": 3, - "1673481600": 4, - "1673568000": 4, - "1673654400": 2, - "1673827200": 3, - "1673913600": 12, - "1674000000": 5, - "1674086400": 4, - "1674172800": 4, - "1674432000": 2, - "1674518400": 3, - "1674604800": 4, - "1674691200": 3, - "1675036800": 5, - "1675123200": 8, - "1675209600": 7, - "1675382400": 3, - "1675641600": 6, - "1675728000": 6, - "1675814400": 16, - "1675900800": 6, - "1675987200": 4, - "1676246400": 8, - "1676332800": 7, - "1676419200": 9, - "1676505600": 20, - "1676592000": 6, - "1676678400": 5, - "1676764800": 5, - "1676851200": 6, - "1676937600": 14, - "1677024000": 3, - "1677110400": 3, - "1677196800": 3, - "1677456000": 3, - "1677542400": 2, - "1677715200": 2, - "1677974400": 2, - "1678060800": 7, - "1678147200": 4, - "1678233600": 2, - "1678492800": 3, - "1678665600": 8, - "1678838400": 2, - "1678924800": 3, - "1679011200": 3, - "1679356800": 4, - "1679443200": 3, - "1679529600": 4, - "1679616000": 2, - "1679961600": 7, - "1680220800": 5, - "1680566400": 2, - "1680652800": 3, - "1681344000": 3, - "1681862400": 4, - "1682121600": 2, - "1682380800": 2, - "1682467200": 3, - "1682553600": 4, - "1683763200": 5, - "1683849600": 9, - "1684195200": 2, - "1684281600": 4, - "1684368000": 5, - "1684454400": 2, - "1684713600": 10, - "1684800000": 9, - "1684972800": 3, - "1685145600": 10, - "1685232000": 3, - "1685318400": 9, - "1685404800": 33, - "1685491200": 15, - "1685577600": 20, - "1685664000": 10, - "1685923200": 4, - "1686096000": 5, - "1686182400": 6, - "1686268800": 2, - "1686614400": 3, - "1687392000": 2, - "1687478400": 3, - "1687737600": 2, - "1687910400": 4, - "1688083200": 2, - "1688774400": 3, - "1689033600": 3, - "1689120000": 2, - "1690156800": 2, - "1693353600": 3, - "1693612800": 2, - "1694563200": 3, - "1695168000": 2, - "1695600000": 2, - "1695859200": 4, - "1696204800": 9, - "1696291200": 8, - "1696377600": 4, - "1696464000": 2, - "1696809600": 7, - "1697500800": 2, - "1698019200": 2, - "1698105600": 6, - "1698278400": 3, - "1698710400": 8, - "1698796800": 4, - "1698883200": 17, - "1698969600": 13, - "1699056000": 3, - "1699228800": 7, - "1699315200": 3, - "1699401600": 9, - "1699488000": 6, - "1699747200": 2, - "1699920000": 3, - "1700006400": 6, - "1700092800": 9, - "1700179200": 2, - "1700265600": 2, - "1700438400": 2, - "1700697600": 2, - "1701043200": 2, - "1701129600": 11, - "1701820800": 2, - "1701907200": 2, - "1702252800": 4, - "1702339200": 20, - "1702512000": 3, - "1702857600": 3, - "1702944000": 6, - "1703030400": 15, - "1703116800": 5, - "1703548800": 3, - "1703635200": 4, - "1703721600": 2, - "1703808000": 11, - "1704067200": 2, - "1704758400": 4, - "1704844800": 8, - "1704931200": 8, - "1705017600": 6, - "1705104000": 4, - "1705276800": 4, - "1705363200": 7, - "1705622400": 2, - "1705881600": 5, - "1705968000": 2, - "1706054400": 4, - "1706140800": 3, - "1706227200": 6, - "1706486400": 2, - "1706572800": 16, - "1706659200": 14, - "1706832000": 3, - "1707004800": 4, - "1707091200": 14, - "1707177600": 5, - "1707264000": 2, - "1707350400": 4, - "1707696000": 6, - "1707782400": 9, - "1707868800": 12, - "1707955200": 8, - "1708041600": 3, - "1708128000": 2, - "1708214400": 2, - "1708300800": 4, - "1708473600": 14, - "1708560000": 26, - "1708646400": 9, - "1708732800": 3, - "1708905600": 4, - "1708992000": 3, - "1709251200": 2, - "1709856000": 2, - "1710374400": 5, - "1711152000": 3, - "1711497600": 3, - "1711584000": 2, - "1711670400": 2, - "1712016000": 3, - "1712102400": 8, - "1712188800": 3, - "1712275200": 4, - "1712448000": 2, - "1712534400": 3, - "1712707200": 2, - "1713139200": 2, - "1713830400": 2, - "1713916800": 2, - "1714003200": 8, - "1714089600": 7, - "1714176000": 3, - "1714348800": 2, - "1715644800": 3, - "1715731200": 4, - "1715817600": 2, - "1715904000": 4, - "1716163200": 12, - "1716249600": 6, - "1716336000": 4, - "1716422400": 7, - "1716508800": 3, - "1716595200": 2, - "1716681600": 2, - "1716768000": 2, - "1717113600": 2, - "1717200000": 2, - "1717977600": 3, - "1718064000": 3, - "1718150400": 5, - "1718236800": 4, - "1718668800": 2, - "1718841600": 3, - "1718928000": 2, - "1719187200": 4, - "1719273600": 2, - "1719532800": 2, - "1719619200": 3, - "1719792000": 2, - "1719878400": 3, - "1719964800": 2, - "1720569600": 2, - "1721692800": 2, - "1721952000": 2, - "1722124800": 2, - "1722384000": 2, - "1723507200": 2, - "1723593600": 6, - "1724025600": 3, - "1724112000": 2, - "1724198400": 2, - "1724716800": 2, - "1724976000": 2, - "1725235200": 5, - "1725321600": 5, - "1725408000": 4, - "1726012800": 2, - "1726185600": 2, - "1726704000": 2, - "1727136000": 2, - "1727222400": 3, - "1727308800": 10, - "1727740800": 3, - "1727827200": 6, - "1727913600": 2, - "1728000000": 2, - "1728086400": 3, - "1728172800": 2, - "1728259200": 4, - "1728432000": 2, - "1728604800": 10, - "1728691200": 6, - "1728777600": 3, - "1728864000": 18, - "1728950400": 7, - "1729036800": 3, - "1729123200": 5, - "1729209600": 7, - "1729468800": 4, - "1729555200": 8, - "1729641600": 12, - "1729728000": 15, - "1729814400": 5, - "1729900800": 3, - "1732060800": 3 - }, - "CSETv1.Autonomy Level": { - "Autonomy1": 847, - "Autonomy3": 370, - "Autonomy2": 261, - "unclear": 104 - }, - "CSETv1.Involving Minor": { - "no": 1469, - "yes": 106, - "maybe": 7 - }, - "CSETv1.Location Region": { - "North America": 781, - "Global": 402, - "Asia": 177, - "Europe": 152, - "Oceania": 64, - "Africa": 5, - "South America": 1 - }, - "CSETv1.Intentional Harm": { - "No. Not intentionally designed to perform harm": 1462, - "Yes. Intentionally designed to perform harm and did create intended harm": 70, - "unclear": 44, - "Yes. Intentionally designed to perform harm but created an unintended harm (a different harm may have occurred)": 2 - }, - "CSETv1.Physical Objects": { - "no": 946, - "yes": 615, - "maybe": 21 - }, - "CSETv1.Rights Violation": { - "no": 1332, - "yes": 168, - "maybe": 82 - }, - "GMF.Known AI Technology": { - "Face Detection": 69, - "Recurrent Neural Network": 55, - "Distributional Learning": 53, - "Convolutional Neural Network": 50, - "Language Modeling": 43, - "Automatic Speech Recognition": 37, - "Acoustic Fingerprint": 35, - "Neural Network": 35, - "Generative Adversarial Network": 33, - "Keyword Filtering": 24, - "Geolocation Data": 15, - "Collaborative Filtering": 14, - "Content-based Filtering": 14, - "Transformer": 14, - "Autoencoder": 13, - "Acoustic Triangulation": 11, - "Character NGrams": 9, - "Classification": 6, - "Gesture Recognition": 6, - "Diverse Data": 4, - "Multimodal Learning": 4, - "Regression": 4, - "Image Segmentation": 2, - "Optical Character Recognition": 1, - "Visual Object Detection": 1 - }, - "CSETv1.Detrimental Content": { - "no": 1281, - "yes": 262, - "maybe": 39 - }, - "CSETv1.AI tools and methods": { - "natural language processing": 235, - "computer vision": 217, - "Natural Language Processesing": 77, - "facial recognition": 55, - "voice recognition": 55, - "neural networks": 54, - "audio transcription": 51, - "machine learning": 51, - "image mapping": 45, - "point mapping": 45, - "unclear": 42, - "object detection": 38, - "human language technology": 32, - "search engine optimization": 31, - "Sensor Data Processing": 29, - "prediction": 28, - "Dijkstra Algorithm": 24, - "face detection": 24, - "logistic regression": 24, - "shortest-path algorithm": 24, - "non-linear SVM": 22, - "signal detection theory": 22, - "facial reconstruction": 21, - "image reconstruction": 21, - "large language models": 17, - "image match": 16, - "text to speech": 16, - "deep learning": 15, - "mesh autoencoders": 13, - "classification": 12, - "vector embedding": 11, - "machine vision": 10, - "large language model": 8, - "regular expressions": 6, - "robotics": 5, - "speech recognition": 5, - "estimation": 4, - "neural network": 3, - "transformer": 3, - "Text to Speach": 2, - "emotion detection": 2, - "large-language models": 2, - "logistic regression model": 2, - "ranking": 2, - "C4.5 algorithm": 1, - "GPT-3": 1, - "Markov decision process": 1, - "SSD network algorithm": 1, - "VCG network": 1, - "YOLO network algorithm": 1, - "classiification": 1, - "convolutional neural networks": 1, - "decision tree": 1, - "linear algebra": 1, - "natural language": 1, - "natural language response": 1, - "pixel embeddings": 1, - "recommendation": 1, - "reinforcement learning": 1, - "response model": 1, - "text scraping": 1, - "unsupervised learning": 1, - "website code": 1 - }, - "CSETv1.Operating Conditions": { - "operationally representative": 63, - "Testing": 29, - "Driverless car": 28, - "night": 25, - "natural disaster - wildfires": 22, - "Twin faces": 21, - "unclear enunciation": 16, - "inclement weather and snow": 6, - "COVID-19 pandemic": 4, - "inclement weather - snow": 2, - "seasonal road closures": 2, - "Non-operational": 1 - }, - "CSETv1.Physical System Type": { - "Manufacturing Robot": 67, - "none": 59, - "rocket/egg shaped, 300 pound, 5 ft. tall security robot": 57, - "Apple iPhone X": 45, - "Amazon Alexa Echo Dot speaker": 35, - "Volvo XC90 SUV": 35, - "Camera": 29, - "Metro train": 29, - "Vehicles (Lexus, Audi, Chrysler Pacifica)": 28, - "Tesla vehicle (Model 3, Model S, Model X)": 25, - "Tesla Model S": 24, - "oval, eight-seater, autonomous, electric shuttle bus": 24, - "Boeing 737 Max airplane": 19, - "Amazon Echo Dot Smart Speaker": 16, - "medical robot": 12, - "Vehicles (Audi Q5 Crossover, Lexus RX400h Crossover)": 11, - "microphones": 11, - "Electric vehicle": 10, - "Tesla Model 3": 6, - "smoke detector": 6, - "A cone-shaped, 400-pound security robot on wheels": 5, - "Nissan vehicles (Rogue, Rogue Hybrid, Rogue Sport, Sentra)": 5, - "Amazon Echo": 4, - "Tesla sedan": 4, - "Vehicles (Volvo, Lexus, Ford, Google, etc.)": 4, - "drone": 4, - "fulfillment robot": 4, - "2016 Tesla Model S": 3, - "Handheld barcode scanner; flat, round mobile robots;": 3, - "4ft tall vaguely humanoid droid on wheels": 2, - "Amazon Echo Dot": 2, - "Humanoid robot": 2, - "body scanner": 2, - "car": 2, - "store security camera": 2, - "Palm VII": 1, - "Starship delivery robot": 1, - "in-van cameras and sensors": 1, - "robotic dog": 1, - "vaguely humanoid robot": 1 - }, - "CSETv1.Sector of Deployment": { - "information and communication": 562, - "transportation and storage": 273, - "Arts, entertainment and recreation": 190, - "wholesale and retail trade": 154, - "administrative and support service activities": 125, - "public administration": 117, - "law enforcement": 114, - "human health and social work activities": 92, - "professional, scientific and technical activities": 71, - "financial and insurance activities": 69, - "manufacturing": 67, - "Education": 41, - "defense": 31, - "accommodation and food service activities": 18, - "other": 10, - "real estate activities": 5, - "other service activities": 2, - "unclear": 2 - }, - "GMF.Potential AI Technology": { - "Regression": 54, - "Image Segmentation": 45, - "Classification": 44, - "Convolutional Neural Network": 44, - "Ensemble Aggregation": 39, - "Distributional Learning": 35, - "3D reconstruction": 29, - "Other domain-specific approaches": 29, - "Face Detection": 27, - "Satellite Imaging": 27, - "Image Classification": 26, - "Intermediate modeling": 26, - "Multimodal Learning": 26, - "Diverse Data": 13, - "Visual Object Detection": 11, - "Clustering": 6, - "Geolocation Data": 4, - "Optical Character Recognition": 4, - "Transformer": 3, - "Acoustic Fingerprint": 2, - "Spectrogram": 2, - "Siamese Network": 1 - }, - "CSETv1.Date of Incident Year": { - "1979": 4, - "1983": 27, - "2000": 13, - "2001": 6, - "2008": 12, - "2009": 28, - "2010": 30, - "2011": 43, - "2012": 41, - "2013": 53, - "2014": 95, - "2015": 78, - "2016": 268, - "2017": 382, - "2018": 181, - "2019": 43, - "2020": 83, - "2021": 82, - "2022": 36, - "2023": 77 - }, - "CSETv1.Entertainment Industry": { - "no": 1401, - "yes": 140, - "maybe": 41 - }, - "CSETv1.Infrastructure Sectors": { - "transportation": 139, - "healthcare and public health": 75, - "financial services": 36, - "defense-industrial base": 31, - "communications": 29, - "information technology": 25, - "emergency services": 23, - "food and agriculture": 8, - "Other": 3, - "commercial facilities": 2, - "government facilities": 2, - "unclear": 2 - }, - "CSETv1.Harm Distribution Basis": { - "none": 1059, - "race": 346, - "sex": 198, - "nation of origin, citizenship, immigrant status": 100, - "religion": 88, - "sexual orientation or gender identity": 73, - "financial means": 44, - "disability": 41, - "geography": 35, - "ideology": 33, - "age": 30, - "familial status (e.g., having or not having children) or pregnancy": 4, - "unclear": 4, - "other": 3 - }, - "CSETv1.Multiple AI Interaction": { - "no": 1525, - "yes": 48, - "maybe": 9 - }, - "GMF.Known AI Technical Failure": { - "Unsafe Exposure or Access": 59, - "Distributional Bias": 55, - "Misuse": 37, - "Adversarial Data": 29, - "Generalization Failure": 28, - "Limited Dataset": 28, - "Black Swan Event": 27, - "Data or Labelling Noise": 27, - "Dataset Imbalance": 27, - "Underfitting": 24, - "Lack of Adversarial Robustness": 23, - "Tuning Issues": 17, - "Inappropriate Training Content": 16, - "Inadequate Anonymization": 13, - "Unauthorized Data": 13, - "Underspecification": 10, - "Context Misidentification": 9, - "Gaming Vulnerability": 5, - "Lack of Capability Control": 5, - "Concept Drift": 4, - "Misinformation Generation Hazard": 3, - "Incomplete Data Attribute Capture": 2, - "Lack of Explainability": 2, - "Lack of Transparency": 2, - "Misconfigured Threshold": 1 - }, - "CSETv1.Clear link to Technology": { - "yes": 1425, - "no": 105, - "maybe": 52 - }, - "CSETv1.Protected Characteristic": { - "no": 1070, - "yes": 505, - "maybe": 7 - }, - "CSETv1.Public Sector Deployment": { - "no": 1324, - "yes": 249, - "maybe": 9 - }, - "CSETv1.Impact on Critical Services": { - "no": 1474, - "yes": 106, - "maybe": 2 - }, - "GMF.Potential AI Technical Failure": { - "Generalization Failure": 135, - "Context Misidentification": 94, - "Dataset Imbalance": 56, - "Underspecification": 52, - "Unauthorized Data": 45, - "Underfitting": 41, - "Hardware Failure": 39, - "Lack of Capability Control": 39, - "Data or Labelling Noise": 38, - "Inadequate Anonymization": 35, - "Overfitting": 34, - "Misinformation Generation Hazard": 33, - "Gaming Vulnerability": 30, - "Distributional Bias": 29, - "Misuse": 29, - "Misconfigured Aggregation": 26, - "Concept Drift": 25, - "Misaligned Objective": 23, - "Misconfigured Threshold": 18, - "Incomplete Data Attribute Capture": 17, - "Limited Dataset": 17, - "Covariate Shift": 16, - "Inadequate Data Sampling": 13, - "Tuning Issues": 10, - "Limited User Access": 6, - "Data Memorization": 5, - "Lack of Explainability": 5, - "Adversarial Data": 4, - "Backup Failure": 4, - "Lack of Transparency": 4, - "Software Bug": 4, - "Robustness Failure": 3, - "Black Box": 2, - "Lack of Adversarial Robustness": 2, - "Problematic Input": 2 - }, - "CSETv1.Location Country (two letters)": { - "US": 769, - "RU": 59, - "CN": 45, - "IN": 41, - "AU": 40, - "IE": 35, - "DE": 32, - "PS": 26, - "VN": 26, - "NZ": 24, - "ID": 19, - "KR": 17, - "GB": 13, - "NL": 6, - "United States": 6, - "CA": 5, - "FR": 5, - "GR": 4, - "LY": 4, - "SE": 4, - "IT": 2, - "JP": 2, - "AR": 1, - "CH": 1, - "IL": 1, - "RS": 1 - }, - "CSETv1.Report, Test, or Study of data": { - "no": 1511, - "yes": 63, - "maybe": 8 - }, - "CSETv1.Special Interest Intangible Harm": { - "no": 862, - "yes": 703, - "maybe": 17 - }, - "CSETv1.User Test in Controlled Conditions": { - "no": 1555, - "yes": 25, - "maybe": 2 - }, - "CSETv1.User Test in Operational Conditions": { - "no": 1367, - "yes": 211, - "maybe": 4 - }, - "CSETv1.Producer Test in Controlled Conditions": { - "no": 1509, - "yes": 69, - "maybe": 4 - }, - "CSETv1.Producer Test in Operational Conditions": { - "no": 1453, - "yes": 129 - }, - "CSETv1.Annotator’s AI special interest intangible harm assessment": { - "no": 949, - "yes": 609, - "maybe": 24 - } + 'Potential AI Technical Failure': ['Tuning Issues', 'Misconfigured Aggregation'], + 'Potential AI Technical Failure Snippets': [ + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"In addition, Kronos is improving a feature meant to help give employees more control over their schedules: Though the software already incorporates employee availability and preferences into its scheduling calculations, improvements to a shift-swapping feature on its employee-facing web and mobile apps will theoretically allow employees to work around conflicts among themselves."', + }, + { + short_name: 'Related Classifications', + value_json: '["Tuning Issues","Misconfigured Aggregation"]', + }, + ], + }, + ], + 'Potential AI Technology': ['Regression', 'Diverse Data'], + 'Potential AI Technology Snippets': [ + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"“You’re waiting on your job to control your life,” she said, with the scheduling software used by her employer dictating everything from “how much sleep Gavin will get to what groceries I’ll be able to buy this month.”"', + }, + { + short_name: 'Related Classifications', + value_json: '["Regression"]', + }, + ], + }, + { + attributes: [ + { + short_name: 'Snippet Text', + value_json: + '"In a follow-up piece, the author, Jodi Kantor, points directly to Kronos\' scheduling software as the root of the problem."', + }, + { + short_name: 'Related Classifications', + value_json: '["Diverse Data"]', + }, + ], + }, + ], + 'Known AI Goal': ['Market Forecasting', 'Scheduling'], + 'Known AI Goal Snippets': [], + }, + 'CSETv1_Annotator-1': { + 'Physical Objects': 'no', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'no', + Deployed: 'yes', + 'Producer Test in Controlled Conditions': 'no', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'no', + 'Tangible Harm': 'tangible harm definitively occurred', + 'AI System': 'yes', + 'AI Harm Level': 'AI tangible harm event', + 'Impact on Critical Services': 'no', + 'Rights Violation': 'no', + 'Involving Minor': 'no', + 'Detrimental Content': 'no', + 'Protected Characteristic': 'no', + 'Harm Distribution Basis': ['none'], + 'Clear link to Technology': 'yes', + 'Annotator’s AI special interest intangible harm assessment': 'no', + 'Sector of Deployment': [ + 'wholesale and retail trade', + 'accommodation and food service activities', + ], + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'Autonomy2', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI tools and methods': '', + 'Special Interest Intangible Harm': 'no', + 'Date of Incident Year': '2014', + 'Multiple AI Interaction': 'no', + Embedded: 'no', + 'Location Country (two letters)': 'US', + 'Location Region': 'North America', + 'Infrastructure Sectors': [], + 'Operating Conditions': '', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': ['schedules', 'worker profiles'], + 'Physical System Type': '', + 'AI Task': ['scheduling', 'productivity optimization'], + }, + CSETv1: { + 'Harm Distribution Basis': ['none'], + 'Sector of Deployment': ['accommodation and food service activities'], + 'Physical Objects': 'no', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'no', + Deployed: 'yes', + 'Producer Test in Controlled Conditions': 'no', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'no', + 'Tangible Harm': 'tangible harm definitively occurred', + 'AI System': 'yes', + 'AI Harm Level': 'AI tangible harm event', + 'Impact on Critical Services': 'no', + 'Rights Violation': 'no', + 'Involving Minor': 'no', + 'Detrimental Content': 'no', + 'Protected Characteristic': 'no', + 'Clear link to Technology': 'yes', + 'Annotator’s AI special interest intangible harm assessment': 'no', + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'Autonomy2', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI tools and methods': '', + 'Special Interest Intangible Harm': 'no', + 'Date of Incident Year': '2014', + 'Multiple AI Interaction': 'no', + Embedded: 'no', + 'Location Country (two letters)': 'US', + 'Location Region': 'North America', + 'Infrastructure Sectors': [], + 'Operating Conditions': '', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': ['schedules', 'worker profiles', 'store traffic'], + 'Physical System Type': '', + 'AI Task': ['scheduling', 'productivity optimization', 'predict store traffic'], + }, + incident_id: 1, + incident_date: '2014-08-14', + epoch_incident_date: 1407974400, + incident_title: + 'Kronos Scheduling Algorithm Allegedly Caused Financial Issues for Starbucks Employees', + incident_description: + 'Kronos’s scheduling algorithm and its use by Starbucks managers allegedly negatively impacted financial and scheduling stability for Starbucks employees, which disadvantaged wage workers.', + objectID: '23', + _snippetResult: { + description: { + value: + "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", + matchLevel: 'none', + }, + text: { + value: + 'For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health', + matchLevel: 'none', + }, + }, + _highlightResult: { + description: { + value: + "Some employees at the coffee chain say it isn't living up to promises to improve the company's labor practices", + matchLevel: 'none', + matchedWords: [], + }, + title: { + value: '​Is Starbucks shortchanging its baristas?', + matchLevel: 'none', + matchedWords: [], + }, + text: { + value: + "For Starbucks (SBUX) barista Kylei Weisse, working at the coffee chain helps him secure health insurance and some extra money while he studies at Georgia Perimeter College. What it doesn't provide is the kind of stable schedule that the company promised its workers last year.\n\n\"It's the wild inconsistency\" of the hours that's a problem, Weisse, 32, said. \"We're supposed to get them 10 days in advance, which often happens, but there's no guarantee. If our manager doesn't get it to us on time, we just have to deal with it.\"\n\nThat became a problem recently when Weisse's manager gave him only a few days notice on his work hours, which ended up conflicting with an anatomy and physiology exam at his college. Weisse ended up paying another worker $20 to take his shift so he could take the exam.\n\nThe short notice is especially frustrating because of Starbucks' vow last year to post employees' schedules at least 10 days in advance, as well as the company's insistence that workers provide at least one-month notice when they need to take a day off.\n\nWhat's behind Starbucks price increases?\n\nWeisse isn't alone in complaining that Starbucks isn't living up to its promises to overhaul its labor practices for its roughly 130,000 baristas. That vow followed an article last year by The New York Times that detailed how workers were struggling to manage childcare and other obligations when the company provided only a few days notice about their schedules.\n\nAbout half of roughly 200 Starbucks baristas in a recent survey said they are still receiving their schedule with less than one week's notice. Others also reported being asked to handle \"clopens,\" split shifts in which employees work a closing shift late into the evening and then an early opening shift the following morning. The company last year promised to end the practice.\n\nOf course, Starbucks isn't alone in using \"just-in-time\" scheduling, with the retail and restaurant industry increasingly turning to software that allows them to change work schedules at the last minute, depending on whether business picks up or slows down. But it is Starbucks that has become a lightning rod on the issue given its vows to improve how it treats employees and its own emphatic claims to valuing workers, whom it labels \"partners.\"\n\n\"Starbucks has the values and wants to do right by their employees,\" said Carrie Gleason, director of the Fair Workweek Initiative at the Center for Popular Democracy, an advocacy group focused on workers'rights, and a co-author of the group's new report on the company's labor practices. \"However, since last year when the company recognized there was a serious problem with the way it scheduled workers and pledged to reform, still so many of the same issues persist.\"\n\nStarbucks didn't respond to requests for comment on the study or on baristas' reports of labor practices that are failing to meet the company's stated goals.\n\nIn an internal memo this week published by Time, Starbucks executive Cliff Burrows wrote that the company couldn't validate the survey, but added that \"the findings suggest, contrary to the expectations we have in place, that some partners are receiving their schedules less than one week in advance and that there is a continuing issue with some partners working a close and then an opening shift the following morning.\" He asks store managers \"to go the extra mile to ensure partners have a consistent schedule.\"\n\nStarbucks ends \"race together\" campaign amid public backlash\n\nTo be sure, some Starbucks workers are receiving at least 10 days notice on their work hours, with the survey finding that about one-third receive two weeks notice and another 18 percent get their schedules three weeks in advance. But that leaves almost half of workers who only receive one week's notice, making it more difficult from them manage other obligations, such as school, family commitments or other jobs.\n\nClopens remain a problem, as well. About 60 percent of workers who have to handle a clopen receive seven or fewer hours of rest between a closing and an opening shift, the study found.\n\nThat's prompted one former Starbucks employee to start a petition to end the practice of scheduling clopens. Ciara Moran noted in her petition that she sometimes was only able to get four or five hours of sleep on the days she was scheduled for clopens. She said she quit her job because she doubted whether it was possible to get ahead given the demands on workers.\n\nEven if Starbucks stuck with its policy of providing eight hours between shifts, that's not enough time, especially given that many workers in the service sector have long commutes, the study said.\n\nAnother issue singled out by the report is Starbucks' practices on sick time. Since paid time off is only available to workers with at least a year on the job, about 40 percent of employees in the survey said they had dealt with barriers in taking sick days.\n\nIn a perfect world, Weisse said he'd like to receive his schedule either a month or a", + matchLevel: 'none', + matchedWords: [], + }, + incident_title: { + value: + 'Kronos Scheduling Algorithm Allegedly Caused Financial Issues for Starbucks Employees', + matchLevel: 'none', + matchedWords: [], + }, + incident_description: { + value: + 'Kronos’s scheduling algorithm and its use by Starbucks managers allegedly negatively impacted financial and scheduling stability for Starbucks employees, which disadvantaged wage workers.', + matchLevel: 'none', + matchedWords: [], + }, + }, + }, + { + authors: ['Marietje Schaake'], + description: + '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ', + epoch_date_downloaded: 1661472000, + epoch_date_modified: 1661731200, + epoch_date_published: 1661385600, + epoch_date_submitted: 1661472000, + image_url: 'https://pbs.twimg.com/media/Fa8rYYVWIAAI0xN?format=jpg', + language: 'en', + report_number: 1967, + source_domain: 'twitter.com', + submitters: ['Cesar Varela'], + title: 'Tweet: @MarietjeSchaake', + name: 'Tweet: @MarietjeSchaake', + url: 'https://twitter.com/MarietjeSchaake/status/1562515297688399873', + tags: [], + editor_notes: '', + cloudinary_id: 'reports/pbs.twimg.com/media/Fa8rYYVWIAAI0xN?format=jpg', + text: '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague\n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden or the Unabomber, but me… How did that happen? What are Meta’s sources?!\n', + mongodb_id: '630c8eb443fe03f46cc8bcc4', + featured: 1, + is_incident_report: true, + is_translated: false, + namespaces: ['CSETv1_Annotator-1', 'CSETv1'], + classifications: [ + 'CSETv1_Annotator-1:Harm Distribution Basis:none', + 'CSETv1_Annotator-1:Sector of Deployment:information and communication', + 'CSETv1_Annotator-1:Physical Objects:no', + 'CSETv1_Annotator-1:Entertainment Industry:no', + 'CSETv1_Annotator-1:Report, Test, or Study of data:no', + 'CSETv1_Annotator-1:Deployed:yes', + 'CSETv1_Annotator-1:Producer Test in Controlled Conditions:no', + 'CSETv1_Annotator-1:Producer Test in Operational Conditions:no', + 'CSETv1_Annotator-1:User Test in Controlled Conditions:no', + 'CSETv1_Annotator-1:User Test in Operational Conditions:yes', + 'CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue', + 'CSETv1_Annotator-1:AI System:yes', + 'CSETv1_Annotator-1:AI Harm Level:none', + 'CSETv1_Annotator-1:Impact on Critical Services:no', + 'CSETv1_Annotator-1:Rights Violation:no', + 'CSETv1_Annotator-1:Involving Minor:no', + 'CSETv1_Annotator-1:Detrimental Content:yes', + 'CSETv1_Annotator-1:Protected Characteristic:no', + 'CSETv1_Annotator-1:Clear link to Technology:yes', + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:yes', + 'CSETv1_Annotator-1:Public Sector Deployment:no', + 'CSETv1_Annotator-1:Autonomy Level:Autonomy1', + 'CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1_Annotator-1:AI tools and methods:large language models', + 'CSETv1_Annotator-1:Special Interest Intangible Harm:yes', + 'CSETv1_Annotator-1:Date of Incident Year:2022', + 'CSETv1_Annotator-1:Multiple AI Interaction:no', + 'CSETv1_Annotator-1:Embedded:no', + 'CSETv1_Annotator-1:Location Region:Global', + 'CSETv1_Annotator-1:Data Inputs:user prompts', + 'CSETv1_Annotator-1:Data Inputs:user queries', + 'CSETv1_Annotator-1:Data Inputs:text', + 'CSETv1_Annotator-1:AI Task:text generation', + 'CSETv1:Harm Distribution Basis:unclear', + 'CSETv1:Sector of Deployment:information and communication', + 'CSETv1:Physical Objects:no', + 'CSETv1:Entertainment Industry:no', + 'CSETv1:Report, Test, or Study of data:no', + 'CSETv1:Deployed:yes', + 'CSETv1:Producer Test in Controlled Conditions:no', + 'CSETv1:Producer Test in Operational Conditions:no', + 'CSETv1:User Test in Controlled Conditions:no', + 'CSETv1:User Test in Operational Conditions:yes', + 'CSETv1:Tangible Harm:no tangible harm, near-miss, or issue', + 'CSETv1:AI System:yes', + 'CSETv1:AI Harm Level:none', + 'CSETv1:Impact on Critical Services:no', + 'CSETv1:Rights Violation:no', + 'CSETv1:Involving Minor:no', + 'CSETv1:Detrimental Content:yes', + 'CSETv1:Protected Characteristic:no', + 'CSETv1:Clear link to Technology:yes', + 'CSETv1:Annotator’s AI special interest intangible harm assessment:yes', + 'CSETv1:Public Sector Deployment:no', + 'CSETv1:Autonomy Level:Autonomy1', + 'CSETv1:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1:AI tools and methods:large language models', + 'CSETv1:Special Interest Intangible Harm:yes', + 'CSETv1:Date of Incident Year:2022', + 'CSETv1:Multiple AI Interaction:no', + 'CSETv1:Embedded:no', + 'CSETv1:Location Region:Global', + 'CSETv1:Data Inputs:user prompts', + 'CSETv1:Data Inputs:user queries', + 'CSETv1:Data Inputs:text', + 'CSETv1:AI Task:text generation', + ], + 'CSETv1_Annotator-1': { + 'Harm Distribution Basis': ['none'], + 'Sector of Deployment': ['information and communication'], + 'Physical Objects': 'no', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'no', + Deployed: 'yes', + 'Producer Test in Controlled Conditions': 'no', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'yes', + 'Tangible Harm': 'no tangible harm, near-miss, or issue', + 'AI System': 'yes', + 'AI Harm Level': 'none', + 'Impact on Critical Services': 'no', + 'Rights Violation': 'no', + 'Involving Minor': 'no', + 'Detrimental Content': 'yes', + 'Protected Characteristic': 'no', + 'Clear link to Technology': 'yes', + 'Annotator’s AI special interest intangible harm assessment': 'yes', + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'Autonomy1', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI tools and methods': ['large language models'], + 'Special Interest Intangible Harm': 'yes', + 'Date of Incident Year': '2022', + 'Multiple AI Interaction': 'no', + Embedded: 'no', + 'Location Country (two letters)': '', + 'Location Region': 'Global', + 'Infrastructure Sectors': [], + 'Operating Conditions': '', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': ['user prompts', 'user queries', 'text'], + 'Physical System Type': '', + 'AI Task': ['text generation'], + }, + CSETv1: { + 'Harm Distribution Basis': ['unclear'], + 'Sector of Deployment': ['information and communication'], + 'Physical Objects': 'no', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'no', + Deployed: 'yes', + 'Producer Test in Controlled Conditions': 'no', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'yes', + 'Tangible Harm': 'no tangible harm, near-miss, or issue', + 'AI System': 'yes', + 'AI Harm Level': 'none', + 'Impact on Critical Services': 'no', + 'Rights Violation': 'no', + 'Involving Minor': 'no', + 'Detrimental Content': 'yes', + 'Protected Characteristic': 'no', + 'Clear link to Technology': 'yes', + 'Annotator’s AI special interest intangible harm assessment': 'yes', + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'Autonomy1', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI tools and methods': ['large language models'], + 'Special Interest Intangible Harm': 'yes', + 'Date of Incident Year': '2022', + 'Multiple AI Interaction': 'no', + Embedded: 'no', + 'Location Country (two letters)': '', + 'Location Region': 'Global', + 'Infrastructure Sectors': [], + 'Operating Conditions': '', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': ['user prompts', 'user queries', 'text'], + 'Physical System Type': '', + 'AI Task': ['text generation'], + }, + incident_id: 2, + incident_date: '2022-08-25', + epoch_incident_date: 1661385600, + incident_title: 'BlenderBot 3 Cited Dutch Politician as a Terrorist', + incident_description: + 'Meta’s conversational AI BlenderBot 3, when prompted “who is a terrorist,“ responded with an incumbent Dutch politician’s name, who was confused about its association.', + objectID: '1967', + _snippetResult: { + description: { + value: + '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ', + matchLevel: 'none', + }, + text: { + value: '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot', + matchLevel: 'none', + }, + }, + _highlightResult: { + description: { + value: + '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague \n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden ', + matchLevel: 'none', + matchedWords: [], + }, + title: { + value: 'Tweet: @MarietjeSchaake', + matchLevel: 'none', + matchedWords: [], + }, + text: { + value: + '😱 Wait! What? Just when you think you’ve seen it all…. Meta’s chatbot replied to the question askedby my colleague\n@kingjen\n: ’Who is a terrorist?’ with my (given) name! That’s right, not Bin Laden or the Unabomber, but me… How did that happen? What are Meta’s sources?!\n', + matchLevel: 'none', + matchedWords: [], + }, + incident_title: { + value: 'BlenderBot 3 Cited Dutch Politician as a Terrorist', + matchLevel: 'none', + matchedWords: [], + }, + incident_description: { + value: + 'Meta’s conversational AI BlenderBot 3, when prompted “who is a terrorist,“ responded with an incumbent Dutch politician’s name, who was confused about its association.', + matchLevel: 'none', + matchedWords: [], + }, + }, + }, + { + authors: ['William Douglas Heaven'], + description: + 'Cuando el covid-19 golpeó Europa en marzo de 2020, los hospitales se sumieron en una crisis sanitaria que aún se entendía mal. “Los médicos realmente no tenían idea de cómo manejar a estos pacientes”, dice Laure Wyn', + epoch_date_downloaded: 1648598400, + epoch_date_modified: 1648598400, + epoch_date_published: 1627603200, + epoch_date_submitted: 1648598400, + image_url: + 'https://wp.technologyreview.com/wp-content/uploads/2021/07/AP_20178810286930.jpg?resize=1200,600', + language: 'en', + report_number: 1551, + source_domain: 'technologyreview.com', + submitters: ['Anonymous'], + title: 'Cientos de herramientas de IA se han creado para detectar el covid. Ninguna de ellas ayudó.', + name: 'Cientos de herramientas de IA se han creado para detectar el covid. Ninguna de ellas ayudó.', + url: 'https://www.technologyreview.com/2021/07/30/1030329/machine-learning-ai-failed-covid-hospital-diagnosis-pandemic/', + tags: [], + editor_notes: '', + cloudinary_id: + 'reports/wp.technologyreview.com/wp-content/uploads/2021/07/AP_20178810286930.jpg?resize=1200,600', + text: 'Cuando el covid-19 golpeó Europa en marzo de 2020, los hospitales se sumieron en una crisis sanitaria que aún se entendía mal. “Los médicos realmente no tenían idea de cómo manejar a estos pacientes”, dice Laure Wynants, una epidemióloga de la Universidad de Maastricht en los Países Bajos, que estudia herramientas predictivas.\n\nPero había datos provenientes de China, que tenía una ventaja de cuatro meses en la carrera para vencer la pandemia. Si los algoritmos de aprendizaje automático pudieran entrenarse con esos datos para ayudar a los médicos a entender lo que estaban viendo y tomar decisiones, podría salvar vidas. “Pensé, `Si hay algún momento en que la IA podría demostrar su utilidad, es ahora`”, dice Wynants. “Tenía mis esperanzas altas.”\n\nNunca sucedió, pero no por falta de esfuerzo. Equipos de investigación de todo el mundo se ofrecieron para ayudar. La comunidad de IA, en particular, se apresuró a desarrollar software que muchos creían permitiría a los hospitales diagnosticar o clasificar a los pacientes más rápido, brindando el tan necesario apoyo a las líneas del frente, en teoría.\n\nAl final, se desarrollaron muchos cientos de herramientas predictivas. Ninguna de ellas hizo una diferencia real, y algunas fueron potencialmente dañinas.\n\nEsa es la condenatoria conclusión de múltiples estudios publicados en los últimos meses. En junio, el Instituto Turing, el centro nacional de ciencia de datos e IA del Reino Unido, publicó un informe que resumía las discusiones en una serie de talleres que realizó a finales de 2020. El claro consenso fue que las herramientas de IA habían tenido poco, si es que algún, impacto en la lucha contra el covid.\n\nNo aptas para uso clínico\n\nEsto refleja los resultados de dos estudios importantes que evaluaron cientos de herramientas predictivas desarrolladas el año pasado. Wynants es la autora principal de uno de ellos, una revisión en el British Medical Journal que aún se está actualizando a medida que se lanzan nuevas herramientas y se prueban las existentes. Ella y sus colegas han examinado 232 algoritmos para diagnosticar pacientes o predecir qué tan enfermos podrían llegar a estar aquellos con la enfermedad. Encontraron que ninguno de ellos era apto para uso clínico. Solo dos han sido señalados como lo suficientemente prometedores para futuras pruebas.\n\n“Es impactante”, dice Wynants. “Entré en esto con algunas preocupaciones, pero esto superó mis temores.”\n\nEl estudio de Wynants está respaldado por otra gran revisión realizada por Derek Driggs, un investigador de aprendizaje automático en la Universidad de Cambridge, y sus colegas, y publicada en Nature Machine Intelligence. Este equipo se centró en modelos de aprendizaje profundo para diagnosticar covid y predecir el riesgo de los pacientes a partir de imágenes médicas, como radiografías de tórax y tomografías computarizadas (TC) de tórax. Examinaron 415 herramientas publicadas y, al igual que Wynants y sus colegas, concluyeron que ninguna era apta para uso clínico.\n\n“Esta pandemia fue una gran prueba para la IA y la medicina”, dice Driggs, quien está trabajando en una herramienta de aprendizaje automático para ayudar a los médicos durante la pandemia. “Habría sido un gran paso para conseguir que el público estuviera de nuestro lado”, dice. “Pero no creo que hayamos pasado esa prueba.”\n\nAmbos equipos encontraron que los investigadores repitieron los mismos errores básicos en la forma en que entrenaron o probaron sus herramientas. Suposiciones incorrectas sobre los datos a menudo significaban que los modelos entrenados no funcionaban como se afirmaba.\n\nWynants y Driggs todavía creen que la IA tiene el potencial de ayudar. Pero les preocupa que pueda ser perjudicial si se construye de la manera incorrecta porque podrían perder diagnósticos o subestimar el riesgo para los pacientes vulnerables. “Hay mucho bombo sobre los modelos de aprendizaje automático y lo que pueden hacer hoy”, dice Driggs.\n\nLas expectativas poco realistas fomentan el uso de estas herramientas antes de que estén listas. Wynants y Driggs dicen que algunos de los algoritmos que examinaron ya se han utilizado en hospitales, y algunos están siendo comercializados por desarrolladores privados. “Temo que puedan haber perjudicado a los pacientes”, dice Wynants.\n\nEntonces, ¿qué salió mal? ¿Y cómo cerramos esa brecha? Si hay un lado positivo, es que la pandemia ha dejado claro para muchos investigadores que la forma en que se construyen las herramientas de IA necesita cambiar. “La pandemia ha puesto en el centro de atención problemas que hemos estado arrastrando durante algún tiempo”, dice Wynants.\n\nQué salió mal\n\nMuchos de los problemas que se descubrieron están relacionados con la mala calidad de los datos que los investigadores utilizaron para desarrollar sus herramientas. La información sobre los pacientes con covid, incluidas las imágenes médicas, se recopiló y compartió en medio de una pandemia global, a menudo por los médicos que luchaban por tratar a esos pacientes. Los investigadores querían ayudar rápidamente, y estos eran los únicos conjuntos de datos públicos disponibles. Pero esto significaba que muchas herramientas se construyeron utilizando datos mal etiquetados o datos de fuentes desconocidas.\n\nDriggs destaca el problema de lo que él llama conjuntos de datos Frankenstein, que se ensamblan a partir de múltiples fuentes y pueden contener duplicados. Esto significa que algunas herramientas terminan siendo probadas con los mismos datos con los que fueron entrenadas, haciéndolas parecer más precisas de lo que son.\n\nTambién ensucia el origen de ciertos conjuntos de datos. Esto puede significar que los investigadores pierdan características importantes que sesgan el entrenamiento de sus modelos. Muchos usaron sin saberlo un conjunto de datos que contenía escaneos de tórax de niños que no tenían covid como sus ejemplos de cómo se veían los casos no covid. Pero como resultado, las IAs aprendieron a identificar niños, no covid.\n\nEl grupo de Driggs entrenó su propio modelo utilizando un conjunto de datos que contenía una mezcla de escaneos tomados cuando los pacientes estaban acostados y de pie. Debido a que los pacientes escaneados mientras estaban acostados tenían más probabilidades de estar gravemente enfermos, la IA aprendió erróneamente a predecir un riesgo grave de covid a partir de la posición de una persona.\n\nEn otros casos, se encontró que algunas IAs estaban detectando la fuente de texto que ciertos hospitales usaban para etiquetar los escaneos. Como resultado, las fuentes de los hospitales con casos más graves se convirtieron en predictores de riesgo de covid.\n\nErrores como estos parecen obvios en retrospectiva. También pueden corregirse ajustando los modelos, si los investigadores son conscientes de ellos. Es posible reconocer las deficiencias y lanzar un modelo menos preciso, pero menos engañoso. Pero muchas herramientas fueron desarrolladas ya sea por investigadores de IA que carecían de la experiencia médica para detectar fallas en los datos o por investigadores médicos que carecían de las habilidades matemáticas para compensar esas fallas.\n\nUn problema más sutil que destaca Driggs es el sesgo de incorporación, o sesgo introducido en el punto en que se etiqueta un conjunto de datos. Por ejemplo, muchas imágenes médicas se etiquetaron según si los radiólogos que las crearon dijeron que mostraban covid. Pero eso incrusta, o incorpora, cualquier sesgo de ese médico en particular en la verdad fundamental de un conjunto de datos. Sería mucho mejor etiquetar una imagen médica con el resultado de una prueba PCR en lugar de la opinión de un médico, dice Driggs. Pero no siempre hay tiempo para sutilezas estadísticas en hospitales ocupados.\n\nEso no ha impedido que algunas de estas herramientas se apresuren a entrar en la práctica clínica. Wynants dice que no está claro cuáles se están utilizando o cómo. Los hospitales a veces dicen que están usando una herramienta solo con fines de investigación, lo que hace difícil evaluar cuánto dependen de ellas los médicos. “Hay mucho secreto”, dice.\n\nWynants le pidió a una empresa que estaba comercializando algoritmos de aprendizaje profundo que compartiera información sobre su enfoque, pero no recibió respuesta. Más tarde encontró varios modelos publicados de investigadores vinculados a esta empresa, todos ellos con un alto riesgo de sesgo. “No sabemos realmente qué implementó la empresa”, dice.\n\nSegún Wynants, algunos hospitales incluso están firmando acuerdos de confidencialidad con proveedores de IA médica. Cuando les preguntó a los médicos qué algoritmos o software estaban usando, a veces le dijeron que no se les permitía decirlo.\n\nCómo solucionarlo\n\n¿Cuál es la solución? Mejores datos ayudarían, pero en tiempos de crisis eso es mucho pedir. Es más importante aprovechar al máximo los conjuntos de datos que tenemos. El movimiento más simple sería que los equipos de IA colaboraran más con los clínicos, dice Driggs. Los investigadores también necesitan compartir sus modelos y divulgar cómo fueron entrenados para que otros puedan probarlos y construir sobre ellos. “Esas son dos cosas que podríamos hacer hoy”, dice. “Y resolverían tal vez', + mongodb_id: '6243a9eedf8b4b62d982817e', + featured: 1, + is_incident_report: true, + is_translated: true, + namespaces: ['CSETv1_Annotator-3', 'CSETv1_Annotator-1'], + classifications: [ + 'CSETv1_Annotator-3:Harm Distribution Basis:unclear', + 'CSETv1_Annotator-3:Sector of Deployment:human health and social work activities', + 'CSETv1_Annotator-3:Physical Objects:yes', + 'CSETv1_Annotator-3:Entertainment Industry:no', + 'CSETv1_Annotator-3:Report, Test, or Study of data:no', + 'CSETv1_Annotator-3:Deployed:yes', + 'CSETv1_Annotator-3:Producer Test in Controlled Conditions:no', + 'CSETv1_Annotator-3:Producer Test in Operational Conditions:no', + 'CSETv1_Annotator-3:User Test in Controlled Conditions:no', + 'CSETv1_Annotator-3:User Test in Operational Conditions:no', + 'CSETv1_Annotator-3:Tangible Harm:unclear', + 'CSETv1_Annotator-3:AI System:yes', + 'CSETv1_Annotator-3:AI Harm Level:none', + 'CSETv1_Annotator-3:Impact on Critical Services:maybe', + 'CSETv1_Annotator-3:Rights Violation:maybe', + 'CSETv1_Annotator-3:Involving Minor:no', + 'CSETv1_Annotator-3:Detrimental Content:no', + 'CSETv1_Annotator-3:Protected Characteristic:maybe', + 'CSETv1_Annotator-3:Clear link to Technology:yes', + 'CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:maybe', + 'CSETv1_Annotator-3:Public Sector Deployment:no', + 'CSETv1_Annotator-3:Autonomy Level:unclear', + 'CSETv1_Annotator-3:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1_Annotator-3:Special Interest Intangible Harm:maybe', + 'CSETv1_Annotator-3:Date of Incident Year:2020', + 'CSETv1_Annotator-3:Multiple AI Interaction:no', + 'CSETv1_Annotator-3:Embedded:maybe', + 'CSETv1_Annotator-3:Location Region:Global', + 'CSETv1_Annotator-3:Infrastructure Sectors:healthcare and public health', + 'CSETv1_Annotator-3:Data Inputs:Unclear', + 'CSETv1_Annotator-1:Physical Objects:no', + 'CSETv1_Annotator-1:Entertainment Industry:no', + 'CSETv1_Annotator-1:Report, Test, or Study of data:maybe', + 'CSETv1_Annotator-1:Deployed:no', + 'CSETv1_Annotator-1:Producer Test in Controlled Conditions:yes', + 'CSETv1_Annotator-1:Producer Test in Operational Conditions:no', + 'CSETv1_Annotator-1:User Test in Controlled Conditions:no', + 'CSETv1_Annotator-1:User Test in Operational Conditions:no', + 'CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue', + 'CSETv1_Annotator-1:AI System:yes', + 'CSETv1_Annotator-1:AI Harm Level:none', + 'CSETv1_Annotator-1:Impact on Critical Services:no', + 'CSETv1_Annotator-1:Rights Violation:no', + 'CSETv1_Annotator-1:Involving Minor:no', + 'CSETv1_Annotator-1:Detrimental Content:no', + 'CSETv1_Annotator-1:Protected Characteristic:no', + 'CSETv1_Annotator-1:Harm Distribution Basis:none', + 'CSETv1_Annotator-1:Special Interest Intangible Harm:no', + 'CSETv1_Annotator-1:Clear link to Technology:maybe', + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no', + 'CSETv1_Annotator-1:Date of Incident Year:2020', + 'CSETv1_Annotator-1:Multiple AI Interaction:no', + 'CSETv1_Annotator-1:Embedded:no', + 'CSETv1_Annotator-1:Location Region:Global', + 'CSETv1_Annotator-1:Data Inputs:patient information', + 'CSETv1_Annotator-1:Data Inputs:medical scans', + 'CSETv1_Annotator-1:Data Inputs:chest scans', + 'CSETv1_Annotator-1:Data Inputs:electronic health records', + 'CSETv1_Annotator-1:Sector of Deployment:human health and social work activities', + 'CSETv1_Annotator-1:Public Sector Deployment:no', + 'CSETv1_Annotator-1:Autonomy Level:Autonomy3', + 'CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm', + 'CSETv1_Annotator-1:AI Task:diagnose patients', + ], + 'CSETv1_Annotator-3': { + 'Harm Distribution Basis': ['unclear'], + 'Sector of Deployment': ['human health and social work activities'], + 'Physical Objects': 'yes', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'no', + Deployed: 'yes', + 'Producer Test in Controlled Conditions': 'no', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'no', + 'Tangible Harm': 'unclear', + 'AI System': 'yes', + 'AI Harm Level': 'none', + 'Impact on Critical Services': 'maybe', + 'Rights Violation': 'maybe', + 'Involving Minor': 'no', + 'Detrimental Content': 'no', + 'Protected Characteristic': 'maybe', + 'Clear link to Technology': 'yes', + 'Annotator’s AI special interest intangible harm assessment': 'maybe', + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'unclear', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI tools and methods': '', + 'Special Interest Intangible Harm': 'maybe', + 'Date of Incident Year': '2020', + 'Multiple AI Interaction': 'no', + Embedded: 'maybe', + 'Location Country (two letters)': '', + 'Location Region': 'Global', + 'Infrastructure Sectors': ['healthcare and public health'], + 'Operating Conditions': '', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': ['Unclear'], + 'Physical System Type': '', + 'AI Task': '', + }, + 'CSETv1_Annotator-1': { + 'Physical Objects': 'no', + 'Entertainment Industry': 'no', + 'Report, Test, or Study of data': 'maybe', + Deployed: 'no', + 'Producer Test in Controlled Conditions': 'yes', + 'Producer Test in Operational Conditions': 'no', + 'User Test in Controlled Conditions': 'no', + 'User Test in Operational Conditions': 'no', + 'Tangible Harm': 'no tangible harm, near-miss, or issue', + 'AI System': 'yes', + 'AI Harm Level': 'none', + 'Impact on Critical Services': 'no', + 'Rights Violation': 'no', + 'Involving Minor': 'no', + 'Detrimental Content': 'no', + 'Protected Characteristic': 'no', + 'Harm Distribution Basis': ['none'], + 'Special Interest Intangible Harm': 'no', + 'Clear link to Technology': 'maybe', + 'Annotator’s AI special interest intangible harm assessment': 'no', + 'Date of Incident Year': '2020', + 'Multiple AI Interaction': 'no', + Embedded: 'no', + 'Location Region': 'Global', + 'Lives Lost': 0, + Injuries: 0, + 'Data Inputs': [ + 'patient information', + 'medical scans', + 'chest scans', + 'electronic health records', + ], + 'Sector of Deployment': ['human health and social work activities'], + 'Public Sector Deployment': 'no', + 'Autonomy Level': 'Autonomy3', + 'Intentional Harm': 'No. Not intentionally designed to perform harm', + 'AI Task': ['diagnose patients'], + }, + incident_id: 3, + incident_date: '2021-07-30', + epoch_incident_date: 1627603200, + incident_title: + 'AI Tools Failed to Sufficiently Predict COVID Patients, Some Potentially Harmful', + incident_description: + 'AI tools failed to sufficiently predict COVID patients, some potentially harmful.', + objectID: '1551', + _snippetResult: { + description: { + value: + 'When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wyn', + matchLevel: 'none', + }, + text: { + value: + 'When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis', + matchLevel: 'none', + }, + }, + _highlightResult: { + description: { + value: + 'When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wyn', + matchLevel: 'none', + matchedWords: [], + }, + title: { + value: 'Hundreds of AI tools have been built to catch covid. None of them helped.', + matchLevel: 'none', + matchedWords: [], }, - "facets_stats": { - "incident_id": { - "min": 1, - "max": 850, - "avg": 336, - "sum": 1304307 - }, - "CSETv1.Injuries": { - "min": 0, - "max": 55000, - "avg": 114, - "sum": 184244 - }, - "CSETv1.Lives Lost": { - "min": 0, - "max": 189, - "avg": 3, - "sum": 5922 - }, - "epoch_incident_date": { - "min": 433382400, - "max": 1732060800, - "avg": 1570344663, - "sum": 6078804192000 - }, - "epoch_date_published": { - "min": 828489600, - "max": 1732147200, - "avg": 1606384020, - "sum": 6218312544000 - }, - "CSETv1.Date of Incident Year": { - "min": 1979, - "max": 2023, - "avg": 2015, - "sum": 3189067 - } + text: { + value: + 'When covid-19 struck Europe in March 2020, hospitals were plunged into a health crisis that was still badly understood. “Doctors really didn’t have a clue how to manage these patients,” says Laure Wynants, an epidemiologist at Maastricht University in the Netherlands, who studies predictive tools.\n\nBut there was data coming out of China, which had a four-month head start in the race to beat the pandemic. If machine-learning algorithms could be trained on that data to help doctors understand what they were seeing and make decisions, it just might save lives. “I thought, ‘If there’s any time that AI could prove its usefulness, it’s now,’” says Wynants. “I had my hopes up.”\n\nIt never happened—but not for lack of effort. Research teams around the world stepped up to help. The AI community, in particular, rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines—in theory.\n\nIn the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful.\n\nThat’s the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK’s national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid.\n\nNot fit for clinical use\n\nThis echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Wynants is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing.\n\n“It’s shocking,” says Wynants. “I went into it with some worries, but this exceeded my fears.”\n\nWynants’s study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use.\n\n“This pandemic was a big test for AI and medicine,” says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. “It would have gone a long way to getting the public on our side,” he says. “But I don’t think we passed that test.”\n\nBoth teams found that researchers repeated the same basic errors in the way they trained or tested their tools. Incorrect assumptions about the data often meant that the trained models did not work as claimed.\n\nWynants and Driggs still believe AI has the potential to help. But they are concerned that it could be harmful if built in the wrong way because they could miss diagnoses or underestimate risk for vulnerable patients. “There is a lot of hype about machine-learning models and what they can do today,” says Driggs.\n\nUnrealistic expectations encourage the use of these tools before they are ready. Wynants and Driggs both say that a few of the algorithms they looked at have already been used in hospitals, and some are being marketed by private developers. “I fear that they may have harmed patients,” says Wynants.\n\nSo what went wrong? And how do we bridge that gap? If there’s an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. “The pandemic has put problems in the spotlight that we’ve been dragging along for some time,” says Wynants.\n\nWhat went wrong\n\nMany of the problems that were uncovered are linked to the poor quality of the data that researchers used to develop their tools. Information about covid patients, including medical scans, was collected and shared in the middle of a global pandemic, often by the doctors struggling to treat those patients. Researchers wanted to help quickly, and these were the only public data sets available. But this meant that many tools were built using mislabeled data or data from unknown sources.\n\nDriggs highlights the problem of what he calls Frankenstein data sets, which are spliced together from multiple sources and can contain duplicates. This means that some tools end up being tested on the same data they were trained on, making them appear more accurate than they are.\n\nIt also muddies the origin of certain data sets. This can mean that researchers miss important features that skew the training of their models. Many unwittingly used a data set that contained chest scans of children who did not have covid as their examples of what non-covid cases looked like. But as a result, the AIs learned to identify kids, not covid.\n\nDriggs’s group trained its own model using a data set that contained a mix of scans taken when patients were lying down and standing up. Because patients scanned while lying down were more likely to be seriously ill, the AI learned wrongly to predict serious covid risk from a person’s position.\n\nIn yet other cases, some AIs were found to be picking up on the text font that certain hospitals used to label the scans. As a result, fonts from hospitals with more serious caseloads became predictors of covid risk.\n\nErrors like these seem obvious in hindsight. They can also be fixed by adjusting the models, if researchers are aware of them. It is possible to acknowledge the shortcomings and release a less accurate, but less misleading model. But many tools were developed either by AI researchers who lacked the medical expertise to spot flaws in the data or by medical researchers who lacked the mathematical skills to compensate for those flaws.\n\nA more subtle problem Driggs highlights is incorporation bias, or bias introduced at the point a data set is labeled. For example, many medical scans were labeled according to whether the radiologists who created them said they showed covid. But that embeds, or incorporates, any biases of that particular doctor into the ground truth of a data set. It would be much better to label a medical scan with the result of a PCR test rather than one doctor’s opinion, says Driggs. But there isn’t always time for statistical niceties in busy hospitals.\n\nThat hasn’t stopped some of these tools from being rushed into clinical practice. Wynants says it isn’t clear which ones are being used or how. Hospitals will sometimes say that they are using a tool only for research purposes, which makes it hard to assess how much doctors are relying on them. “There’s a lot of secrecy,” she says.\n\nWynants asked one company that was marketing deep-learning algorithms to share information about its approach but did not hear back. She later found several published models from researchers tied to this company, all of them with a high risk of bias. “We don’t actually know what the company implemented,” she says.\n\nAccording to Wynants, some hospitals are even signing nondisclosure agreements with medical AI vendors. When she asked doctors what algorithms or software they were using, they sometimes told her they weren’t allowed to say.\n\nHow to fix it\n\nWhat’s the fix? Better data would help, but in times of crisis that’s a big ask. It’s more important to make the most of the data sets we have. The simplest move would be for AI teams to collaborate more with clinicians, says Driggs. Researchers also need to share their models and disclose how they were trained so that others can test them and build on them. “Those are two things we could do today,” he says. “And they would solve mayb', + matchLevel: 'none', + matchedWords: [], }, - "exhaustiveFacetsCount": true, - "exhaustiveFacetValues": false, - "exhaustiveNbHits": true, - "exhaustive": { - "facetsCount": true, - "facetValues": false, - "nbHits": true + incident_title: { + value: + 'AI Tools Failed to Sufficiently Predict COVID Patients, Some Potentially Harmful', + matchLevel: 'none', + matchedWords: [], }, - "query": "", - "params": "distinct=false&facetFilters=%5B%5B%22is_incident_report%3Atrue%22%5D%5D&facets=%5B%22CSETv1.AI%20Harm%20Level%22%2C%22CSETv1.AI%20System%22%2C%22CSETv1.AI%20System%20Description%22%2C%22CSETv1.AI%20Tangible%20Harm%20Level%20Notes%22%2C%22CSETv1.AI%20Task%22%2C%22CSETv1.AI%20tools%20and%20methods%22%2C%22CSETv1.Annotation%20Status%22%2C%22CSETv1.Annotator%22%2C%22CSETv1.Annotator%E2%80%99s%20AI%20special%20interest%20intangible%20harm%20assessment%22%2C%22CSETv1.Autonomy%20Level%22%2C%22CSETv1.Clear%20link%20to%20Technology%22%2C%22CSETv1.Clear%20link%20to%20technology%22%2C%22CSETv1.Data%20Inputs%22%2C%22CSETv1.Date%20of%20Incident%20Day%22%2C%22CSETv1.Date%20of%20Incident%20Month%22%2C%22CSETv1.Date%20of%20Incident%20Year%22%2C%22CSETv1.Deployed%22%2C%22CSETv1.Detrimental%20Content%22%2C%22CSETv1.Embedded%22%2C%22CSETv1.Entertainment%20Industry%22%2C%22CSETv1.Entities%22%2C%22CSETv1.Estimated%20Date%22%2C%22CSETv1.Estimated%20Harm%20Quantities%22%2C%22CSETv1.Harm%20Distribution%20Basis%22%2C%22CSETv1.Harm%20Domain%22%2C%22CSETv1.Harmed%20Class%20of%20Entities%22%2C%22CSETv1.Impact%20on%20Critical%20Services%22%2C%22CSETv1.Incident%20Number%22%2C%22CSETv1.Infrastructure%20Sectors%22%2C%22CSETv1.Injuries%22%2C%22CSETv1.Intentional%20Harm%22%2C%22CSETv1.Involving%20Minor%22%2C%22CSETv1.Lives%20Lost%22%2C%22CSETv1.Location%20City%22%2C%22CSETv1.Location%20Country%20(two%20letters)%22%2C%22CSETv1.Location%20Region%22%2C%22CSETv1.Location%20State%2FProvince%20(two%20letters)%22%2C%22CSETv1.Multiple%20AI%20Interaction%22%2C%22CSETv1.Notes%20(%20Tangible%20Harm%20Quantities%20Information)%22%2C%22CSETv1.Notes%20(AI%20Functionality%20and%20Techniques)%22%2C%22CSETv1.Notes%20(AI%20special%20interest%20intangible%20harm)%22%2C%22CSETv1.Notes%20(Environmental%20and%20Temporal%20Characteristics)%22%2C%22CSETv1.Notes%20(Information%20about%20AI%20System)%22%2C%22CSETv1.Notes%20(special%20interest%20intangible%20harm)%22%2C%22CSETv1.Operating%20Conditions%22%2C%22CSETv1.Peer%20Reviewer%22%2C%22CSETv1.Physical%20Objects%22%2C%22CSETv1.Physical%20System%20Type%22%2C%22CSETv1.Producer%20Test%20in%20Controlled%20Conditions%22%2C%22CSETv1.Producer%20Test%20in%20Operational%20Conditions%22%2C%22CSETv1.Protected%20Characteristic%22%2C%22CSETv1.Public%20Sector%20Deployment%22%2C%22CSETv1.Quality%20Control%22%2C%22CSETv1.Report%2C%20Test%2C%20or%20Study%20of%20data%22%2C%22CSETv1.Rights%20Violation%22%2C%22CSETv1.Sector%20of%20Deployment%22%2C%22CSETv1.Special%20Interest%20Intangible%20Harm%22%2C%22CSETv1.Tangible%20Harm%22%2C%22CSETv1.There%20is%20a%20potentially%20identifiable%20specific%20entity%20that%20experienced%20the%20harm%22%2C%22CSETv1.User%20Test%20in%20Controlled%20Conditions%22%2C%22CSETv1.User%20Test%20in%20Operational%20Conditions%22%2C%22GMF.Known%20AI%20Goal%22%2C%22GMF.Known%20AI%20Goal%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Goal%20Snippets%22%2C%22GMF.Known%20AI%20Technical%20Failure%22%2C%22GMF.Known%20AI%20Technical%20Failure%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Technical%20Failure%20Snippets%22%2C%22GMF.Known%20AI%20Technology%22%2C%22GMF.Known%20AI%20Technology%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Technology%20Snippets%22%2C%22GMF.Potential%20AI%20Goal%22%2C%22GMF.Potential%20AI%20Goal%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Goal%20Snippets%22%2C%22GMF.Potential%20AI%20Technical%20Failure%22%2C%22GMF.Potential%20AI%20Technical%20Failure%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Technical%20Failure%20Snippets%22%2C%22GMF.Potential%20AI%20Technology%22%2C%22GMF.Potential%20AI%20Technology%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Technology%20Snippets%22%2C%22authors%22%2C%22classifications%22%2C%22epoch_date_published%22%2C%22epoch_incident_date%22%2C%22flag%22%2C%22incident_id%22%2C%22is_incident_report%22%2C%22language%22%2C%22namespaces%22%2C%22source_domain%22%2C%22submitters%22%2C%22tags%22%5D&highlightPostTag=__%2Fais-highlight__&highlightPreTag=__ais-highlight__&hitsPerPage=28&maxValuesPerFacet=999&page=0&query=&tagFilters=", - "index": "instant_search-en-featured", - "renderingContent": {}, - "processingTimeMS": 3, - "processingTimingsMS": { - "_request": { - "roundTrip": 160 - }, - "afterFetch": { - "format": { - "highlighting": 3, - "snippeting": 6, - "total": 12 - }, - "total": 2 - }, - "fetch": { - "total": 1 - }, - "total": 3 + incident_description: { + value: + 'AI tools failed to sufficiently predict COVID patients, some potentially harmful.', + matchLevel: 'none', + matchedWords: [], }, - "serverTimeMS": 16, - } - ] -}; \ No newline at end of file + }, + }, + ], + nbHits: 3, + page: 0, + nbPages: 1, + hitsPerPage: 28, + facets: { + flag: { + true: 36, + }, + tags: { + response: 163, + deepfake: 23, + 'generative AI': 12, + CSAM: 5, + journalism: 5, + Amazon: 4, + minors: 4, + 'racial bias': 4, + Audio: 3, + Deepfakes: 3, + Grok: 3, + Scam: 3, + TikTok: 3, + 'deepfake porn': 3, + fraud: 3, + 'social media': 3, + Alexa: 2, + Celebrities: 2, + Children: 2, + 'Data privacy': 2, + Education: 2, + Extortion: 2, + 'Facial recognition': 2, + 'Kamala Harris': 2, + Misinformation: 2, + Nazi: 2, + Nudify: 2, + Privacy: 2, + 'algorithmic bias': 2, + chatbot: 2, + 'child abuse': 2, + 'child sex abuse material': 2, + 'content moderation': 2, + data: 2, + 'driverless cars': 2, + elections: 2, + 'facial recognition technology': 2, + 'health insurance': 2, + 'lack of consent': 2, + products: 2, + '#response': 1, + '2FA': 1, + 'AI Music': 1, + 'AI bias': 1, + 'AI content farms': 1, + 'AI facial feature modification': 1, + 'AI image alteration': 1, + 'AI meal planner': 1, + 'AI misinterpretation of professionalism': 1, + 'AI racial representation': 1, + 'AI research': 1, + 'AI-authored books': 1, + 'AI-generated news sites': 1, + Advertisement: 1, + Alaska: 1, + 'Amazon Photos': 1, + 'Apple Photos': 1, + Army: 1, + 'Barack Obama': 1, + 'British politics': 1, + 'California law': 1, + 'Central America': 1, + ChatGPT: 1, + Chatbots: 1, + Cheating: 1, + 'Child sexual abuse material': 1, + 'Clearview AI': 1, + 'Criminal activity': 1, + Crypto: 1, + DPA: 1, + 'Dark Web': 1, + 'Data exposure': 1, + 'Data retention': 1, + 'Data set': 1, + 'Donald Trump': 1, + 'Ed Tech': 1, + 'Educational AI': 1, + 'Election interference': 1, + 'Elon Musk': 1, + 'Encouragement of violence': 1, + Exam: 1, + FTC: 1, + 'Fake news': 1, + 'False positive': 1, + Figma: 1, + GitHub: 1, + 'Google Photos': 1, + IWF: 1, + 'Image recognition': 1, + India: 1, + 'Internet Watch Foundation': 1, + 'LAION-5B': 1, + 'LGBTQ+': 1, + 'Latin America': 1, + Lawsuit: 1, + 'Mental illness': 1, + 'Microsoft Azure': 1, + 'Microsoft OneDrive': 1, + Music: 1, + Nudification: 1, + OSHA: 1, + Photoshop: 1, + 'Plagiarism detector': 1, + 'Political manipulation': 1, + Porn: 1, + 'Queer representation': 1, + Robots: 1, + Russia: 1, + 'SAS tokens': 1, + 'Security vulnerability': 1, + Slovakia: 1, + 'South America': 1, + 'South Korea': 1, + Spain: 1, + Spam: 1, + 'Stable Diffusion': 1, + Tesla: 1, + 'Training data': 1, + Turkish: 1, + Twitter: 1, + Undress: 1, + Video: 1, + 'Workplace accident': 1, + X: 1, + accountability: 1, + art: 1, + 'automated decision-making': 1, + autopilot: 1, + 'biometric data': 1, + celebrity: 1, + 'child exploitation': 1, + 'child pornography': 1, + comedy: 1, + commercials: 1, + 'conference call': 1, + 'copyright infringement': 1, + court: 1, + 'coverage denial': 1, + 'customer service': 1, + 'data accuracy': 1, + 'data integrity': 1, + 'deepfake audio': 1, + 'deepfaked audio': 1, + 'digital crime': 1, + disinformation: 1, + 'editing error': 1, + 'false arrest': 1, + 'false citations': 1, + 'gender bias': 1, + 'generated books': 1, + geopolitics: 1, + 'google ai search': 1, + hallucination: 1, + 'hazardous advice': 1, + 'healthcare algorithms': 1, + 'healthcare modeling': 1, + 'healthcare technology': 1, + 'inappropriate content': 1, + 'influence campaigns': 1, + 'journalism ethics': 1, + 'labor relations': 1, + 'language model malfunction': 1, + law: 1, + legal: 1, + 'legal dispute': 1, + loans: 1, + 'medical claims': 1, + 'medical diagnosis': 1, + misidentification: 1, + 'news aggregation': 1, + 'news network': 1, + 'offensive content': 1, + 'official report using false generated AI': 1, + pedestrians: 1, + 'photo editing': 1, + 'political disinformation': 1, + politics: 1, + 'posthumous use': 1, + retail: 1, + robotics: 1, + 'self-driving': 1, + 'synthetic imagery': 1, + 'targeted advertising': 1, + 'text-to-image generation': 1, + underwriting: 1, + 'website poll': 1, + 'wifi connectivity': 1, + 'workplace death': 1, + 'wrongful arrest': 1, + }, + authors: { + 'Associated Press': 25, + Reuters: 23, + 'BBC News': 22, + 'James Vincent': 22, + 'Kashmir Hill': 18, + 'Christopher Knaus': 13, + 'Katyanna Quach': 13, + 'Benj Edwards': 12, + 'Noor Al-Sibai': 12, + 'Alex Hern': 11, + 'Andrew J. Hawkins': 11, + 'Fred Lambert': 11, + 'Tom Simonite': 11, + 'Emma Roth': 10, + 'Maggie Harrison': 10, + 'Will Oremus': 10, + 'Chloe Xiang': 9, + 'Pranshu Verma': 9, + 'Sam Levin': 9, + 'Surya Mattu': 9, + Bloomberg: 8, + 'Dan Milmo': 8, + 'David Shepardson': 8, + 'Drew Harwell': 8, + 'James Whitbrook': 8, + 'Julia Angwin': 8, + 'Karen Hao': 8, + 'The Guardian': 8, + 'Tom Krisher': 8, + 'Aarian Marshall': 7, + 'Ashley Belanger': 7, + 'Jon Fingas': 7, + 'Kat Tenbarge': 7, + 'Mack DeGeurin': 7, + 'Matthew Gault': 7, + 'Pesala Bandara': 7, + 'Samantha Cole': 7, + 'Steven Lee Myers': 7, + 'Agence France-Presse': 6, + 'Gerrit De Vynck': 6, + 'Jon Christian': 6, + 'Kathleen McGrory': 6, + 'Matt McFarland': 6, + 'Neal E. Boudette': 6, + 'Olivia Solon': 6, + 'Ryan Mac': 6, + 'Samuel Gibbs': 6, + 'Sky News': 6, + 'Tim Cushing': 6, + 'Todd Feathers': 6, + 'Will Knight': 6, + 'Brandon Vigliarolo': 5, + 'Britney Nguyen': 5, + 'Cade Metz': 5, + 'Dan Goodin': 5, + 'Faiz Siddiqui': 5, + 'Garance Burke': 5, + 'Human Rights Watch': 5, + 'James Titcomb': 5, + 'Jason Koebler': 5, + 'Jay Peters': 5, + 'John Goreham': 5, + 'Josh Taylor': 5, + 'Julie Jargon': 5, + 'Lora Kolodny': 5, + 'Melissa Heikkilä': 5, + 'Natasha Lomas': 5, + 'Natasha Singer': 5, + 'Nico Grant': 5, + 'Nitasha Tiku': 5, + 'Ryan Felton': 5, + 'Shannon Bond': 5, + 'Stuart A. Thompson': 5, + 'The Economist': 5, + 'Thomas Claburn': 5, + 'Tiffany Hsu': 5, + 'Timothy B. Lee': 5, + 'Trevor Mogg': 5, + 'Victor Tangermann': 5, + 'Wikipedia Editors': 5, + Afp: 4, + 'Arden Dier': 4, + BBC: 4, + 'Benjamin Weiser': 4, + 'Brad Templeton': 4, + 'CBS News': 4, + 'Casey Ross': 4, + 'Cyrus Farivar': 4, + 'Dave Lee': 4, + 'David Meyer': 4, + 'Derek B. Johnson': 4, + 'Elizabeth Stanton': 4, + 'Federal Trade Commission': 4, + 'Gabriel Geiger': 4, + 'Glenn Mcdonald': 4, + 'Hilke Schellmann': 4, + 'IFL Science': 4, + 'Isobel Asher Hamilton': 4, + 'Jack Morse': 4, + 'James Cook': 4, + 'Jeff Larson': 4, + 'Johana Bhuiyan': 4, + 'Jon Brodkin': 4, + 'Jordan Pearson': 4, + 'Joseph Cox': 4, + 'Kevin Roose': 4, + 'Khari Johnson': 4, + 'Kyle Wiggers': 4, + 'Lauren Kirchner': 4, + 'Lauren Leffer': 4, + 'Lucas Ropek': 4, + 'Luke Henriques-Gomes': 4, + 'Madison Malone Kircher': 4, + 'Maggie Harrison Dupré': 4, + 'Matt Novak': 4, + 'National Highway Traffic Safety Administration': 4, + 'Privacy International': 4, + RT: 4, + 'Rachel Metz': 4, + 'Rob Price': 4, + 'Sean Keach': 4, + 'Shannon Liao': 4, + 'Srishti Deoras': 4, + 'The Inquirer': 4, + 'Thomas Orsolya': 4, + 'Wes Davis': 4, + 'Aaron Rieke': 3, + 'Abigail Beall': 3, + 'Adam Satariano': 3, + 'Adrienne Roberts': 3, + 'Aj Dellinger': 3, + 'Al Jazeera': 3, + 'Alan Levin': 3, + 'Aleksandra Korolova': 3, + 'Alex Kasprak': 3, + 'Alex Seitz-Wald': 3, + 'Ananya Bhattacharya': 3, + 'Andrew Higgins': 3, + 'Annie Palmer': 3, + 'Ariel Zilber': 3, + 'Australian Broadcasting Corporation': 3, + 'Bill McCarthy': 3, + 'Blake Brittain': 3, + 'Bobby Allyn': 3, + 'Brendan Pierson': 3, + 'Brett Molina': 3, + 'Brian Barrett': 3, + 'Cara McGoogan': 3, + 'Carl Franzen': 3, + 'Caroline Haskins': 3, + 'Catherine Thorbecke': 3, + 'Cecily Mauran': 3, + 'Charles Pulliam-Moore': 3, + 'Chris Burt': 3, + 'City News Service': 3, + 'Cory Doctorow': 3, + 'Dan Levine': 3, + 'Daniel Wu': 3, + 'Dara Kerr': 3, + 'Davey Alba': 3, + 'David Gilbert': 3, + 'David Grossman': 3, + 'David Lumb': 3, + 'Department of Motor Vehicles': 3, + 'Devin Coldewey': 3, + 'Dhruv Mehrotra': 3, + 'Douglas Heaven': 3, + 'Douglas Macmillan': 3, + 'Duncan Riley': 3, + 'Eileen Guo': 3, + 'Elizabeth Dwoskin': 3, + 'Elizabeth Napolitano': 3, + 'Emilia David': 3, + 'Erielle Delzer': 3, + 'Faustine Ngila': 3, + Foxglove: 3, + 'Frank Landymore': 3, + 'Gareth Corfield': 3, + 'Geoffrey A. Fowler': 3, + 'Gianluca Mezzofiore': 3, + 'Global Witness': 3, + 'Heather Somerville': 3, + 'Heather Vogell': 3, + 'Hyunjoo Jin': 3, + IANS: 3, + 'Issie Lapowsky': 3, + 'Jack Gillum': 3, + 'Jackie Snow': 3, + 'Jane Wakefield': 3, + 'Jeff Horwitz': 3, + 'Jefferson Graham': 3, + 'Jeffrey Dastin': 3, + 'Jim Waterson': 3, + 'Jonathan M. Gitlin': 3, + 'Kevin Okemwa': 3, + 'Kieren McCarthy': 3, + 'Kit Eaton': 3, + 'Kris Holt': 3, + 'Kyle Orland': 3, + 'Lauren Kaori Gurley': 3, + 'Leon Yin': 3, + 'Louise Matsakis': 3, + 'Loukia Papadopoulos': 3, + 'Luke Dormehl': 3, + 'Mallory Locklear': 3, + 'Marco Marcelline': 3, + 'Matt Binder': 3, + "Matt O'Brien": 3, + 'Megan Cerullo': 3, + 'Mia Sato': 3, + 'Michael Kan': 3, + 'Mikael Thalen': 3, + 'Mike Murphy': 3, + 'Morgan Meaker': 3, + 'Naomi Nix': 3, + 'National Transportation Safety Board': 3, + 'Natt Garun': 3, + 'Neil Bedi': 3, + 'News Desk': 3, + 'Nicholas Thompson': 3, + 'Nick Robins-Early': 3, + 'Nicolas Kayser-Bril': 3, + OpenAI: 3, + Pandaily: 3, + 'Paris Martineau': 3, + 'Patrick Collinson': 3, + 'Paul Mozur': 3, + 'Phoebe Weston': 3, + 'Phys Org': 3, + 'Reuters Editorial': 3, + 'Rob Beschizza': 3, + 'Rob Stumpf': 3, + 'Robert Booth': 3, + 'Ronan Glon': 3, + 'Ryan Whitwam': 3, + 'Sam Biddle': 3, + 'Saqib Shah': 3, + 'Sejal Sharma': 3, + 'Sheera Frenkel': 3, + 'Stephen Edelstein': 3, + Synced: 3, + 'Team Latestly': 3, + 'The Straits Times': 3, + 'Thomas Barrabi': 3, + 'Thomas Brewster': 3, + 'Thomas Germain': 3, + 'Thomas Macaulay': 3, + 'Tim De Chant': 3, + 'Times of India': 3, + 'Tom Warren': 3, + 'Trisha Thadani': 3, + 'Tristan Greene': 3, + 'ABP News Bureau': 2, + 'Aaron Glantz': 2, + 'Aaron Horowitz': 2, + 'Aaron Mok': 2, + 'Aaron Sankin': 2, + 'Aaron Schaffer': 2, + 'Abby Ohlheiser': 2, + 'Abhirup Roy': 2, + 'Adam Clark Estes': 2, + 'Adi Robertson': 2, + 'Adrianna Nine': 2, + 'Adrianne Jeffries': 2, + 'Adrienne Lafrance': 2, + 'Aimee Picchi': 2, + 'Alan Martin': 2, + 'Alan Mislove': 2, + 'Alex Davies': 2, + 'Alex Pigman': 2, + 'Alfred Ng': 2, + 'Alisha Rahaman Sarkar': 2, + 'Alistair Barr': 2, + 'Alistair Charlton': 2, + 'Allana Akhtar': 2, + 'Alyssa Mercante': 2, + 'Amanda Hoover': 2, + 'Amnesty International': 2, + 'Amrita Khalid': 2, + 'Ana Gutiérrez': 2, + 'Andrew Griffin': 2, + 'Andrew Hobbs': 2, + 'Andrew Liptak': 2, + 'Andy Greenberg': 2, + 'Angela Yang': 2, + 'Anjana Samant': 2, + 'Anne Hayes': 2, + 'Annie Gilbertson': 2, + Anonymous: 2, + 'Anthony Cuthbertson': 2, + 'Antoine Allen': 2, + 'Aol Staff': 2, + 'App Drivers and Couriers Union': 2, + 'April Glaser': 2, + 'April Rubin': 2, + 'Ariana Tobin': 2, + 'Ariel Bogle': 2, + 'Ariel Herbert-Voss': 2, + 'Arijeta Lajka': 2, + 'Arvind Narayanan': 2, + 'Arwa Mahdawi': 2, + 'Avi Asher-Schapiro': 2, + 'Ayang Macdonald': 2, + 'Barbara Ortutay': 2, + 'Beatrice Nolan': 2, + 'Bella Cacciatore': 2, + 'Ben Dickson': 2, + 'Ben Ellery': 2, + 'Beth Mole': 2, + 'Bill Donahue': 2, + 'Billy Perrigo': 2, + 'Blake Montgomery': 2, + 'Bob Herman': 2, + 'Bradford Betz': 2, + 'Brian Bushard': 2, + 'Brian Fung': 2, + 'Buster Hein': 2, + 'CNN Business': 2, + 'CNN Newsource': 2, + 'Cailin Loesch': 2, + 'Caitlin Kelly': 2, + 'Carlos E. Perez': 2, + 'Caroline Mimbs Nyce': 2, + "Caroline O'Donovan": 2, + 'Casey Tolan': 2, + "Cathy O'Neil": 2, + 'Chaim Gartenberg': 2, + 'Charles Duhigg': 2, + 'Charlie Warzel': 2, + 'Chitra Ramaswamy': 2, + 'Choe Sang-Hun': 2, + 'Chris Isidore': 2, + 'Chris Matyszczyk': 2, + 'Chris Morris': 2, + 'Chris Stokel-Walker': 2, + 'Chris Vallance': 2, + 'Christian De Looper': 2, + 'Christopher Rosa': 2, + 'Chuck Dinerstein': 2, + 'Chuong Nguyen': 2, + 'Ciaran Lyons': 2, + 'Clare Duffy': 2, + 'Colin Lecher': 2, + 'Conor Cawley': 2, + 'Dan Robitzski': 2, + 'Daniel Croft': 2, + 'Daniel E. Ho': 2, + 'Danny Palmer': 2, + 'David Koenig': 2, + 'David Z. Morris': 2, + 'De Elizabeth': 2, + 'Debra Cassens Weiss': 2, + 'Dell Cameron': 2, + 'Department of Justice': 2, + "Donie O'Sullivan": 2, + 'Dustin Volz': 2, + 'Ed Yong': 2, + 'Edward Helmore': 2, + 'Elizabeth Culliford': 2, + 'Emily Dreyfuss': 2, + 'Emily Flitter': 2, + 'Emmanuelle Saliba': 2, + 'Eric Wallace': 2, + 'Erin McCormick': 2, + 'Ethan Baron': 2, + 'Florian Tramèr': 2, + 'France 24': 2, + 'Garrett M. Graff': 2, + 'Gary Craig': 2, + 'Gene Maddaus': 2, + 'George Joseph': 2, + 'Georgia Wells': 2, + 'Giulia Carbonaro': 2, + 'Globe Staff': 2, + 'Grégoire Sauvage': 2, + 'Guardian staff and agencies': 2, + 'Gustavo Henrique Ruffo': 2, + 'Hamza Shaban': 2, + 'Hannah Devlin': 2, + 'Hannah Getahun': 2, + 'Hannah Knowles': 2, + 'Hany Farid': 2, + 'Harry Suhartono': 2, + 'Heather Chen': 2, + 'Heather Stewart': 2, + 'Henry Belot': 2, + 'Hope Reese': 2, + 'Igor Bonifacic': 2, + 'India Today Web Desk': 2, + 'Indo-Asian News Service': 2, + 'Insurance Journal': 2, + Inverse: 2, + 'Isaac Chotiner': 2, + 'Isaiah Richard': 2, + 'Itech Post': 2, + 'Ivan Mehta': 2, + 'Jack Brewster': 2, + 'Jack Smith Iv': 2, + 'Jackie Lieberman': 2, + 'Jacob Serebrin': 2, + 'Jacob Snow': 2, + 'Jake Offenhartz': 2, + 'Jake Peterson': 2, + 'James Clayton': 2, + 'James Felton': 2, + 'James Purtill': 2, + 'James Zou': 2, + 'Janus Rose': 2, + 'Jarni Blakkarly': 2, + 'Jazper Lu': 2, + 'Jean Mackenzie': 2, + 'Jeff Kao': 2, + 'Jennifer Cowan': 2, + 'Jennifer L. Doleac': 2, + 'Jennifer Langston': 2, + 'Jeremy B. Merrill': 2, + 'Jeremy Kahn': 2, + 'Jess Weatherbed': 2, + 'Jessica Guynn': 2, + 'Joan Lowy': 2, + 'Jodi Kantor': 2, + 'Joe Kukura': 2, + 'Joe Patrice': 2, + 'Joe Vaccarelli': 2, + 'Joel Hruska': 2, + 'John Glenday': 2, + 'John Pring': 2, + 'John Simerman': 2, + 'John West': 2, + 'Jonathan Limehouse': 2, + 'Jonathan Vanian': 2, + 'Joseph Hincks': 2, + 'Joseph Menn': 2, + 'Joshua Bote': 2, + 'Joshua Thurston': 2, + 'Julia Carrie Wong': 2, + 'Julian E. Barnes': 2, + 'Julie Johnsson': 2, + 'Justin Pritchard': 2, + 'Jyoti Mann': 2, + 'Kalhan Rosenblatt': 2, + 'Karin Kovary Solymos': 2, + 'Kate Proctor': 2, + 'Kath Xu': 2, + 'Katherine Lee': 2, + 'Kathleen Magramo': 2, + 'Katie Notopoulos': 2, + 'Keith Wagstaff': 2, + 'Kelvin Chan': 2, + 'Keoni Everington': 2, + 'Khamila Mulia': 2, + 'Kristian Lum': 2, + 'Kyle Barr': 2, + 'Kylie Cheung': 2, + 'Lara Pearce': 2, + 'Larry Alton': 2, + 'Latanya Sweeney': 2, + 'Lauren Rhue': 2, + 'Lauren Sforza': 2, + 'Leena Nasir': 2, + 'Leonardo Nicoletti': 2, + 'Lewin Day': 2, + 'Li Tao': 2, + 'Liam Reilly': 2, + 'Lily Hay Newman': 2, + 'Linette Lopez': 2, + 'Lorenzo Arvanitis': 2, + 'Lucas Manfredi': 2, + 'Lydia Horne': 2, + 'Maia Szalavitz': 2, + 'Marco Della Cava': 2, + 'Marie Solis': 2, + 'Mariella Moon': 2, + 'Marissa Gerchick': 2, + 'Mark Frauenfelder': 2, + 'Mark Sweney': 2, + 'Mary Papenfuss': 2, + 'Matt Burgess': 2, + 'Matt Simon': 2, + 'Matt Wille': 2, + 'Matthew Butterick': 2, + 'Matthew Guariglia': 2, + 'Matthew Humphries': 2, + 'Matthew Jagielski': 2, + 'Matthew Phelan': 2, + 'Matthias Bastian': 2, + 'Maya Oppenheim': 2, + 'McKenzie Sadeghi': 2, + 'Megan Farokhmanesh': 2, + 'Megan Hickey': 2, + 'Melia Robinson': 2, + 'Melissa del Bosque': 2, + 'Michael L. Diamond': 2, + 'Michael Savage': 2, + 'Michael Tarm': 2, + 'Michelle Butterfield': 2, + 'Mick Akers': 2, + Microsoft: 2, + 'Mike Masnick': 2, + 'Mike Melanson': 2, + 'Mike Spector': 2, + 'Mirna Alsharif': 2, + 'Mitchell Clark': 2, + 'Moinak Pal': 2, + 'Molly Liebergall': 2, + 'Moohita Kaur Garg': 2, + 'Morgan Sung': 2, + 'Muhammad Ali': 2, + 'Nadine Freischlad': 2, + 'Nam Hyun-woo': 2, + "Natalie O'Neill": 2, + 'Natasha Duarte': 2, + 'Nathaniel Gleicher': 2, + 'Neil Vigdor': 2, + 'Niamh Ancell': 2, + 'Nicholas Diakopoulos': 2, + 'Nicholas Lezard': 2, + 'Nick Evershed': 2, + 'Nick Statt': 2, + 'Nicole Clark': 2, + 'Nikki Williams': 2, + 'Nilay Patel': 2, + 'Nilesh Christopher': 2, + 'Niniek Karmini': 2, + 'Noam Shemtov': 2, + 'Onur Demirkol': 2, + 'Paige Skinner': 2, + 'Paul Egan': 2, + 'Paul Lukas': 2, + 'Pete Bigelow': 2, + 'Philip Marcelo': 2, + 'Phillip Tracy': 2, + 'Piotr Sapiezynski': 2, + 'Pocharapon Neammanee': 2, + 'Political Reporter': 2, + 'RT Staff': 2, + 'Rachel Kraus': 2, + 'Rachel Lerman': 2, + 'Ramishah Maruf': 2, + 'Rasha Ali': 2, + 'Raymond Wong': 2, + 'Rebecca Hill': 2, + 'Rene Marsh': 2, + 'Rita Liao': 2, + 'Rob Thubron': 2, + 'Rob Waugh': 2, + 'Romy Ellenbogen': 2, + 'Russ Mitchell': 2, + 'Ryan Daws': 2, + 'Ryan McNeal': 2, + 'Sally Ho': 2, + 'Sapna Maheshwari': 2, + "Sara Ashley O'Brien": 2, + 'Sarah Frier': 2, + 'Sarah Perez': 2, + 'Sayash Kapoor': 2, + 'Scott Nover': 2, + 'Sean B. McGregor': 2, + 'Sean McGregor': 2, + 'Sean Szymkowski': 2, + 'Sebastien Bell': 2, + 'Shannon Thaler': 2, + 'Shanti Das': 2, + 'Sharon Adarlo': 2, + 'Sharon Goldman': 2, + 'Shira Ovide': 2, + 'Shivali Best': 2, + 'Shona Ghosh': 2, + 'Sigal Samuel': 2, + 'Siladitya Ray': 2, + 'Simon Willison': 2, + 'Soo Youn': 2, + 'Sophia Waterfield': 2, + 'Sophie Beiers': 2, + 'Sophie Nieto-Munoz': 2, + Staff: 2, + 'Stan Schroeder': 2, + 'Steph Maj Swanson': 2, + 'Stephanie Wykstra': 2, + 'Stephen Kafeero': 2, + 'Steve Crowe': 2, + 'Steve Stecklow': 2, + 'Steven Loveday': 2, + 'Steven Musil': 2, + Stuff: 2, + 'Tarak Shah': 2, + 'Tate Ryan-Mosley': 2, + 'Tatum Hunter': 2, + 'Taylor Donovan Barnett': 2, + 'Tech Transparency Project': 2, + 'The Express Tribune': 2, + 'The Hindu Bureau': 2, + 'The Independent': 2, + 'Tim McNicholas': 2, + 'Tobi Jegede': 2, + 'Tony Ho Tran': 2, + 'Tony Kovaleski': 2, + 'Tony Tran': 2, + 'Turkish Minute': 2, + 'Varsha Bansal': 2, + 'Victoria Woollaston': 2, + 'Vittoria Elliott': 2, + 'Wall Street Journal': 2, + 'Will Bedingfield': 2, + 'Will McCurdy': 2, + 'William Gavin': 2, + 'William Hoffman': 2, + 'William Isaac': 2, + 'Yiwen Lu': 2, + 'Yuval Abraham': 2, + 'bne IntelliNews': 2, + 천호성: 2, + '@BeauHD': 1, + '@DearKick': 1, + '@FSD_in_6m': 1, + '@Robert_AIZI': 1, + '@Waqas': 1, + '@an_open_mind': 1, + '@kawazacky': 1, + '@msravi': 1, + '@ohgustie': 1, + 'A.J. Perez': 1, + AAP: 1, + 'ABC 7 Eyewitness News': 1, + 'ABC Australia': 1, + 'AI Addict': 1, + 'AP/Reuters': 1, + 'Aaditi Lele': 1, + 'Aakar Patel': 1, + 'Aamir Sheikh': 1, + 'Aaron Blake': 1, + 'Aaron Gordon': 1, + 'Aaron Gregg': 1, + 'Aaron Holmes': 1, + 'Aaron Huff': 1, + 'Aaron Katersky': 1, + 'Aaron Krolik': 1, + 'Aaron Mak': 1, + 'Aaron Mamiit': 1, + 'Aaron Moss': 1, + 'Aaron Smith': 1, + 'Aaron Tilley': 1, + 'Aashish Kumar Shrivastava': 1, + 'Aatif Sulleyman': 1, + 'Abbie Richards': 1, + 'Abby Monteil': 1, + 'Abdul Moeed': 1, + Abdullah: 1, + 'Abeba Birhane': 1, + 'Abigail Rubenstein': 1, + 'Abrar Al-Heeti': 1, + 'Abubakar Abid': 1, + 'Abubakar Idris': 1, + 'Adam Carlson': 1, + 'Adam Downer': 1, + 'Adam Garfinkle': 1, + 'Adam Hadhazy': 1, + 'Adam Healy': 1, + 'Adam Kalai': 1, + 'Adam Lam': 1, + 'Adam Rawnsley': 1, + 'Adam Roberts': 1, + 'Adam Rowe': 1, + 'Aditi Sen': 1, + 'Aditya Bahl': 1, + 'Aditya Kalra': 1, + 'Adrian Hopgood': 1, + 'Adrian Morrow': 1, + 'Adrian Weckler': 1, + 'Adriana Lee': 1, + 'Agence France Presse': 1, + 'Aicha El Hammar Castano': 1, + 'Aiden Pink': 1, + 'Aimee Lucido': 1, + 'Aimee Ortiz': 1, + 'Aisha Malik': 1, + 'Akash Pandey': 1, + 'Aki Anastasiou': 1, + 'Aki Peritz': 1, + 'Al-Monitor Staff': 1, + 'Alain Sherter': 1, + 'Alan Boyle': 1, + 'Alan Friedman': 1, + 'Alastair Gale': 1, + 'Albert Tait': 1, + 'Alberto Luperon': 1, + 'Alberto Romero': 1, + 'Aleks Phillips': 1, + 'Alessandra Kellermann': 1, + 'Alessandro Mascellino': 1, + 'Alex Arger': 1, + 'Alex Brenninkmeijer': 1, + 'Alex Cadier': 1, + 'Alex Carr': 1, + 'Alex Clark': 1, + 'Alex Farber': 1, + 'Alex Heath': 1, + 'Alex Isenstadt': 1, + 'Alex Johnson': 1, + 'Alex Kaplan': 1, + 'Alex Kierstein': 1, + 'Alex Lockie': 1, + 'Alex Mitchell': 1, + 'Alex Moersen': 1, + 'Alex Portée': 1, + 'Alex Tsiaoussidis': 1, + 'Alex Wawro': 1, + 'Alexa Cimino': 1, + 'Alexa Corse': 1, + 'Alexa Hagerty': 1, + 'Alexander Hanff': 1, + 'Alexander J Martin': 1, + 'Alexander Martin': 1, + 'Alexander Puutio': 1, + 'Alexandra Blogier': 1, + 'Alexandra Kravariti': 1, + 'Alexandra Marquez': 1, + 'Alexandra Miller': 1, + 'Alexandra S. Levine': 1, + 'Alexei Alexis': 1, + 'Alexis Haden': 1, + 'Alexis Kleinman': 1, + 'Alexis Madrigal': 1, + 'Ali A. Jessani': 1, + 'Ali Breland': 1, + 'Ali Swenson': 1, + 'Alice Hearing': 1, + 'Alice Kohn': 1, + 'Alice Milliken': 1, + 'Alice Richardson': 1, + 'Alice Wright': 1, + 'Alicia Valenski': 1, + 'Alina Oprea': 1, + 'Alisha Green': 1, + 'Alison Cutler': 1, + 'Alison Sider': 1, + 'Alix Langone': 1, + 'Alizeh Kohari': 1, + 'Allison Koenecke': 1, + 'Allison Levitsky': 1, + 'Alma Fabiani': 1, + 'Alvin E. Roth': 1, + 'Alyce McFadden': 1, + 'Alyse Stanley': 1, + 'Alyssa Braithwaite': 1, + 'Alyssa Goard': 1, + 'Alyssa Guzman': 1, + 'Alyssa Newcomb': 1, + 'Alyssa Rinelli': 1, + 'Amal Attar-Guzman': 1, + 'Amanda Claypool': 1, + 'Amanda Cochran': 1, + 'Amanda Geffner': 1, + 'Amanda Kooser': 1, + 'Amanda Lewis': 1, + 'Amanda Schupak': 1, + 'Amanda Shaw': 1, + 'Amaya Ross': 1, + 'Amelia Butterly': 1, + 'Amelia Gentleman': 1, + 'Amelia Lucas': 1, + 'Amelia McGuire': 1, + 'Amelia Mcdonell-Parry': 1, + 'Amiah Taylor': 1, + 'Amit Katwala': 1, + 'Amiya Moretta': 1, + 'Amrutha Pagad': 1, + 'Amy Beth Hanson': 1, + 'Amy Gardner': 1, + 'Amy Kraft': 1, + 'Amy Martinez': 1, + 'Ana María Arévalo Gosen': 1, + 'Anagha Srikanth': 1, + 'Anand Tamboli': 1, + 'Ananya Gairola': 1, + 'Andrea Bernstein': 1, + 'Andrea Blanco': 1, + 'Andreo Calonzo': 1, + 'Andrew Buncombe': 1, + 'Andrew Court': 1, + 'Andrew Deck': 1, + 'Andrew Desiderio': 1, + 'Andrew Guthrie Ferguson': 1, + 'Andrew Hutchinson': 1, + 'Andrew Longhi': 1, + 'Andrew Nam': 1, + 'Andrew Orlowski': 1, + 'Andrew Papachristos': 1, + 'Andrew Schlaikjer': 1, + 'Andrew Smith': 1, + 'Andrew Tangel': 1, + 'Andrew Thompson': 1, + 'Andrew Trotman': 1, + 'Andrew Van Dam': 1, + 'Andrew Wilks': 1, + 'Andy Brown': 1, + 'Andy Miller': 1, + 'Andy Mukherjee': 1, + 'Andy Nghiem': 1, + 'Andy Pasztor': 1, + 'Andy Verity': 1, + 'Angel Saunders': 1, + 'Angela B.': 1, + 'Angela Lai': 1, + 'Angelica Mari': 1, + 'Angry Asian Man': 1, + 'Angus Morrison': 1, + Ani: 1, + 'Ani Nenkova': 1, + 'Anisa Harrasy': 1, + 'Anjana Nair': 1, + 'Ankita Garg': 1, + 'Anna Andrews': 1, + 'Anna Bahney': 1, + 'Anna Houlahan': 1, + 'Anna Iovine': 1, + 'Anna Kutz': 1, + 'Anna Loup': 1, + 'Anna Lucente Sterling': 1, + 'Anna Tong': 1, + 'Annalyn Kurtz': 1, + 'Anne Marie Lee': 1, + 'Anneli L. Tostar': 1, + 'Annie Minoff': 1, + 'Annie Palmer For Dailymail.Com': 1, + 'Anthony G. Attrino': 1, + 'Antoine Gara': 1, + 'Anton Ekker': 1, + 'Antonio Andrés-Pueyo': 1, + 'Antonio Madeira': 1, + 'Antonio Pequeño IV': 1, + 'Antonio Villas-Boas': 1, + 'Anubhav Sharma': 1, + 'Anumita Kaur': 1, + 'Anupam Dikhit': 1, + 'Anupama Airy': 1, + 'Anurag Baruah': 1, + 'Anushe Fawaz': 1, + 'Anushree Hede': 1, + 'Aparna Iyer': 1, + 'Apple On Amazon': 1, + 'Archie Hamilton': 1, + 'Archie Mitchell': 1, + 'Ardi Wirdana': 1, + 'Arfa Javaid': 1, + 'Ariana Baio': 1, + 'Arianna Evers': 1, + 'Aric Jenkins': 1, + 'Arielle Pardes': 1, + 'Arif Perdana': 1, + 'Aris Folley': 1, + 'Arjun Kharpal': 1, + 'Armin Ronacher': 1, + 'Armin Rosen': 1, + 'Artemis Moshtaghian': 1, + 'Arthur Holland Michel': 1, + 'Arun Ramesh': 1, + 'Aryan Prakash': 1, + 'Aryn Plax': 1, + 'Ashleigh Hollowell': 1, + 'Ashley Bardhan': 1, + 'Ashley Capoot': 1, + 'Ashley Collman': 1, + 'Ashley Halsey III': 1, + 'Ashley Hume': 1, + 'Ashley May': 1, + 'Ashley McBride': 1, + 'Ashley Rodriguez': 1, + 'Ashlie D. Stevens': 1, + 'Ashok Ramprasad': 1, + 'Ashutosh Tripathi': 1, + 'Asia Grace': 1, + Aspyhackr: 1, + 'Atlanta Journal-Constitution': 1, + 'Audie Cornish': 1, + 'Audrey Ash': 1, + 'Augustine Fou': 1, + 'Austin Mullen': 1, + 'Australian Associated Press': 1, + 'Ava Mutchler': 1, + Avaaz: 1, + 'Averi Kremposky': 1, + 'Avinash A': 1, + 'Aye Min Thant': 1, + 'Aylin Caliskan': 1, + 'Aylin Elci': 1, + 'Ayushman Kaul': 1, + 'B J Robertson': 1, + 'BR Data': 1, + 'Ban Barkawi': 1, + 'Barbara Liston': 1, + 'Barbara S. Peterson': 1, + 'Barbara Zmušková': 1, + 'Barry Sookman': 1, + 'Basileal Imana': 1, + 'Bayu Anggorojati': 1, + 'Bbc Trending': 1, + 'Beatrice Verhoeven': 1, + 'Becca Smouse': 1, + 'Becky Bargh': 1, + Beebom: 1, + 'Beijing Newsroom': 1, + 'Bel Air Association': 1, + 'Belga News Agency': 1, + 'Belinda Grant Geary': 1, + 'Ben Beaumont-Thomas': 1, + 'Ben Brumfield': 1, + 'Ben Chu': 1, + 'Ben Cohen': 1, + 'Ben Cost': 1, + 'Ben Dubow': 1, + 'Ben Eltham': 1, + 'Ben Geman': 1, + 'Ben Guarino': 1, + 'Ben Kaufman': 1, + 'Ben Lovejoy': 1, + 'Ben Nimmo': 1, + 'Ben Popper': 1, + 'Ben Sales': 1, + 'Ben Strauss': 1, + 'Ben Sullivan': 1, + 'Ben Travis': 1, + 'Ben Warwick': 1, + 'Ben Woods': 1, + 'Ben Yelin': 1, + 'Benedict Smith': 1, + 'Benita Kolovos': 1, + 'Benjamin Fearnow': 1, + 'Benjamin Goggin': 1, + 'Benjamin Haas': 1, + 'Benjamin Lindsay': 1, + 'Benjamin Perkin': 1, + 'Benjamin Schneider': 1, + 'Benjamin Snyder': 1, + 'Benjamin Wallace-Wells': 1, + 'Berenice Baker': 1, + 'Berhan Taye': 1, + 'Bernie Woodall': 1, + 'Beth Greenfield': 1, + 'Bethan McKernan': 1, + 'Bethan Sexton': 1, + 'Bethany Hallam': 1, + 'Bethany White': 1, + 'Bettina Büchel': 1, + 'Bev Stephans': 1, + 'Bharat Sharma': 1, + 'Bhavya Sukheja': 1, + 'Bianca Bosker': 1, + 'Bianca Britton': 1, + 'Bigger Law Firm Magazine': 1, + 'Bijan Stephen': 1, + 'Bill Cash': 1, + 'Bill Christensen': 1, + 'Bill Snyder': 1, + 'Biman Mukherji': 1, + 'Bindu Bansinath': 1, + 'Biometrics Commissioner': 1, + 'Björn ten Seldam': 1, + 'Black Voices Editor': 1, + 'Blair Hanley Frank': 1, + 'Blake Harper': 1, + 'Blake Lew-Merwin': 1, + 'Bobbie Johnson': 1, + 'Bogdan Popa': 1, + 'Boston Herald': 1, + 'Botego Inc': 1, + 'Brad Anderson': 1, + 'Braden Bjella': 1, + 'Bradley Hope': 1, + 'Bradley Jolly': 1, + 'Brandon Turkus': 1, + 'Brayden Lindrea': 1, + 'Breanna Barraclough': 1, + 'Bree Fowler': 1, + 'Breeze Liu': 1, + 'Brenda Medina': 1, + 'Brenden Gallagher': 1, + 'Brett Wilkins': 1, + 'Brian Fraga': 1, + 'Brian Lisi': 1, + 'Brian Melley': 1, + 'Brian Merchant': 1, + 'Brian Resnick': 1, + 'Brian Welk': 1, + 'Brian Whitton': 1, + 'Brianna Sacks': 1, + 'Bridget Chavez': 1, + 'Bridget McArthur': 1, + 'Brie Barbee': 1, + 'Brie Stimson': 1, + 'Brieanna J Frank': 1, + 'Brinda A. Thomas': 1, + 'Brittny Mejia': 1, + 'Bruce Brown': 1, + 'Bruce Handy': 1, + 'Bruce Schneier': 1, + 'Bryan Clark': 1, + 'Bryan Logan': 1, + 'Bálint Miklós': 1, + }, + language: { + en: 3838, + de: 6, + es: 5, + fr: 4, + it: 4, + th: 4, + vi: 4, + ko: 2, + hi: 1, + nl: 1, + pt: 1, + 'zh-CN': 1, + }, + namespaces: { + 'CSETv1_Annotator-1': 1898, + CSETv1: 1617, + CSETv0: 1217, + 'CSETv1_Annotator-2': 954, + 'CSETv1_Annotator-3': 436, + GMF: 402, + }, + submitters: { + 'Daniel Atherton': 1349, + Anonymous: 862, + 'Khoa Lam': 416, + 'Roman Yampolskiy': 372, + 'Catherine Olsson': 208, + 'Kate Perkins': 139, + 'Ingrid Dickinson (CSET)': 73, + 'Thomas Giallella (CSET)': 68, + 'Sean McGregor': 42, + 'Patrick Hall': 33, + 'Srishti Khemka (CSET)': 33, + 'Roman Lutz': 26, + 'Devon Colmer (CSET)': 22, + 'CSET annotators': 21, + 'Sonali Pednekar (CSET)': 21, + 'Logan B': 20, + 'Collin Starkweather': 15, + 'Janet Schwartz': 14, + 'Luna McNulty': 11, + 'Kevin Paeth': 10, + 'Nickie Demakos': 9, + 'Andrew Hundt': 7, + 'Frank Guo': 7, + 'Ayrton San Joaquin': 6, + 'Neama Dadkhahnikoo': 6, + 'Arthit Suriyawongkul': 5, + 'Andrew Gamino-Cheong': 4, + 'Michael Simon (ForHumanity)': 4, + 'Viviana Rivera': 4, + 'Catharina Doria': 3, + 'Cesar Varela': 3, + 'Helen Zhu': 3, + 'Sundar Narayanan (ForHumanity)': 3, + 'Alice RT (ForHumanity)': 2, + 'Charlie Pownall': 2, + 'Effy Elden': 2, + "Fionntan O'Donnell": 2, + 'Irina Borisova King': 2, + 'Maud Stiernet (ForHumanity)': 2, + 'Nathan Butters': 2, + 'Nik Martelaro': 2, + 'Octavia Occident': 2, + 'Sundar Narayanan': 2, + 'William Schindhelm Georg': 2, + kepa: 2, + reubot: 2, + '21five': 1, + 'Alexis Monks': 1, + 'Alice Villano': 1, + 'Antonio Buffelli (ForHumanity)': 1, + 'Ashley Casovan': 1, + 'Beatrice Moissinac': 1, + 'Carol Anderson (ForHumanity)': 1, + 'Charlie Wang': 1, + 'Christopher Maratos': 1, + 'Corey Abshire': 1, + 'Eric Horvitz': 1, + 'Fabio Xie': 1, + 'Fion Lee-Madan (Fairly AI)': 1, + 'Gerry Chng': 1, + 'Inbal - For Humanity': 1, + 'Inbal - ForHumanity': 1, + Isabelle: 1, + 'Joanna (ForHumanity)': 1, + 'Jodi Masters-Gonzales': 1, + 'Johann Hanssen (ForHumanity)': 1, + 'Joshua Poore': 1, + 'K M': 1, + 'Karson Elmgren': 1, + 'Leon Overweel': 1, + 'Lilianna Smith': 1, + 'Lilly Ryan': 1, + 'Madison Malone Kircher': 1, + 'Matteo Dora': 1, + 'Mitt Regan Jr. (BABL AI)': 1, + 'Natalia DLC': 1, + 'Nga Than': 1, + 'Nick Stockton': 1, + 'Parul Pandey': 1, + 'Ryan Carrier (ForHumanity)': 1, + 'Samuel Curtis': 1, + 'Scott Cambo': 1, + 'Subhabrata Majumdar': 1, + 'Subho Majumdar': 1, + TNT: 1, + 'Tuoi Tran': 1, + 'Wiebke Hutiri': 1, + 'Yukti Handa': 1, + kepae: 1, + }, + incident_id: { + '1': 14, + '2': 17, + '3': 19, + '4': 25, + '5': 12, + '6': 28, + '7': 6, + '8': 10, + '9': 7, + '10': 10, + '11': 15, + '12': 1, + '13': 9, + '14': 7, + '15': 24, + '16': 24, + '17': 22, + '18': 11, + '19': 27, + '20': 22, + '22': 22, + '23': 24, + '24': 27, + '25': 11, + '26': 24, + '27': 27, + '28': 30, + '29': 2, + '30': 28, + '31': 29, + '32': 21, + '33': 4, + '34': 35, + '35': 20, + '36': 25, + '37': 33, + '38': 11, + '39': 29, + '40': 22, + '41': 28, + '42': 1, + '43': 4, + '44': 1, + '45': 29, + '46': 6, + '47': 9, + '48': 22, + '49': 10, + '50': 24, + '51': 27, + '52': 29, + '53': 18, + '54': 13, + '55': 16, + '56': 7, + '57': 39, + '58': 5, + '59': 10, + '60': 23, + '61': 1, + '63': 1, + '64': 1, + '65': 1, + '66': 16, + '67': 24, + '68': 30, + '69': 12, + '70': 4, + '71': 28, + '72': 26, + '73': 8, + '74': 11, + '75': 1, + '76': 1, + '77': 5, + '78': 1, + '79': 3, + '80': 2, + '81': 1, + '82': 1, + '83': 1, + '84': 1, + '86': 2, + '87': 1, + '88': 2, + '89': 1, + '91': 5, + '92': 6, + '93': 4, + '94': 2, + '95': 4, + '96': 1, + '97': 1, + '98': 1, + '99': 1, + '100': 1, + '101': 6, + '102': 2, + '103': 5, + '104': 1, + '105': 1, + '106': 13, + '107': 2, + '108': 3, + '109': 1, + '110': 1, + '111': 5, + '112': 11, + '113': 1, + '114': 1, + '115': 3, + '116': 2, + '117': 4, + '118': 3, + '119': 4, + '120': 1, + '121': 4, + '122': 1, + '123': 4, + '124': 7, + '125': 3, + '126': 4, + '127': 12, + '128': 2, + '129': 1, + '131': 2, + '132': 1, + '133': 1, + '134': 2, + '135': 2, + '136': 1, + '137': 1, + '138': 6, + '139': 2, + '140': 1, + '141': 2, + '142': 1, + '143': 1, + '144': 6, + '145': 3, + '146': 3, + '147': 2, + '148': 1, + '149': 4, + '150': 3, + '151': 10, + '152': 2, + '153': 3, + '154': 1, + '155': 2, + '156': 2, + '157': 1, + '158': 1, + '160': 2, + '161': 3, + '162': 1, + '163': 2, + '164': 1, + '165': 2, + '166': 2, + '167': 1, + '168': 2, + '169': 5, + '170': 3, + '171': 1, + '172': 3, + '173': 1, + '174': 4, + '175': 5, + '176': 1, + '177': 5, + '178': 8, + '179': 3, + '180': 3, + '181': 2, + '182': 2, + '183': 6, + '184': 3, + '185': 4, + '186': 7, + '187': 3, + '188': 4, + '189': 6, + '190': 4, + '191': 2, + '192': 2, + '193': 1, + '194': 1, + '195': 12, + '196': 5, + '197': 4, + '198': 9, + '199': 7, + '200': 1, + '201': 2, + '202': 5, + '203': 3, + '204': 4, + '205': 4, + '206': 4, + '207': 4, + '208': 8, + '209': 5, + '210': 3, + '211': 7, + '212': 4, + '213': 5, + '214': 1, + '215': 1, + '216': 5, + '217': 2, + '218': 3, + '219': 1, + '220': 4, + '221': 2, + '222': 1, + '223': 1, + '224': 1, + '225': 2, + '226': 3, + '227': 3, + '228': 1, + '229': 2, + '230': 1, + '231': 4, + '232': 4, + '233': 1, + '234': 2, + '235': 1, + '236': 1, + '238': 1, + '239': 1, + '240': 5, + '241': 5, + '242': 1, + '243': 2, + '244': 1, + '245': 1, + '246': 2, + '248': 2, + '249': 2, + '250': 1, + '251': 3, + '252': 1, + '253': 4, + '254': 2, + '255': 9, + '256': 1, + '257': 4, + '258': 2, + '259': 2, + '260': 2, + '261': 8, + '262': 4, + '263': 1, + '264': 1, + '265': 2, + '266': 8, + '267': 10, + '268': 4, + '270': 1, + '271': 3, + '272': 3, + '273': 1, + '274': 2, + '275': 2, + '276': 1, + '277': 1, + '278': 3, + '279': 3, + '280': 2, + '281': 3, + '282': 3, + '283': 1, + '284': 6, + '285': 1, + '286': 2, + '288': 4, + '289': 2, + '290': 3, + '291': 6, + '292': 3, + '293': 7, + '294': 4, + '295': 5, + '296': 3, + '297': 3, + '298': 1, + '299': 1, + '300': 2, + '301': 1, + '302': 1, + '303': 2, + '304': 1, + '305': 2, + '306': 3, + '307': 1, + '308': 2, + '309': 4, + '310': 8, + '311': 2, + '312': 4, + '313': 2, + '314': 1, + '315': 1, + '316': 1, + '317': 1, + '318': 2, + '319': 5, + '320': 4, + '321': 14, + '322': 2, + '323': 4, + '324': 6, + '325': 2, + '326': 3, + '327': 1, + '328': 2, + '329': 1, + '330': 1, + '331': 2, + '332': 4, + '333': 6, + '334': 2, + '335': 8, + '336': 4, + '337': 9, + '339': 14, + '340': 1, + '341': 5, + '343': 2, + '344': 1, + '345': 1, + '346': 5, + '347': 3, + '348': 3, + '349': 2, + '350': 2, + '351': 1, + '352': 4, + '353': 4, + '354': 5, + '355': 4, + '356': 2, + '357': 3, + '358': 1, + '359': 1, + '360': 3, + '361': 1, + '362': 1, + '363': 1, + '364': 1, + '366': 1, + '367': 1, + '368': 10, + '369': 1, + '370': 1, + '371': 3, + '372': 3, + '373': 12, + '374': 8, + '375': 3, + '376': 5, + '377': 1, + '378': 2, + '379': 2, + '380': 5, + '381': 1, + '382': 1, + '383': 2, + '384': 2, + '385': 6, + '386': 3, + '387': 1, + '388': 1, + '389': 2, + '390': 1, + '391': 2, + '392': 2, + '393': 1, + '394': 2, + '395': 4, + '396': 1, + '397': 2, + '398': 3, + '399': 3, + '400': 1, + '401': 4, + '402': 1, + '403': 2, + '404': 2, + '405': 2, + '406': 1, + '407': 1, + '408': 1, + '409': 3, + '410': 1, + '411': 1, + '412': 4, + '413': 2, + '414': 1, + '415': 5, + '416': 3, + '417': 4, + '418': 3, + '419': 3, + '420': 11, + '421': 9, + '422': 1, + '423': 5, + '424': 4, + '425': 2, + '426': 1, + '427': 3, + '428': 3, + '429': 4, + '430': 21, + '431': 3, + '432': 1, + '433': 10, + '434': 8, + '435': 4, + '436': 9, + '437': 4, + '438': 3, + '439': 3, + '440': 6, + '441': 5, + '443': 25, + '444': 1, + '445': 4, + '446': 5, + '447': 2, + '448': 1, + '449': 4, + '450': 8, + '451': 5, + '452': 2, + '453': 1, + '454': 2, + '455': 7, + '456': 7, + '457': 3, + '458': 1, + '459': 2, + '460': 2, + '461': 4, + '462': 5, + '463': 3, + '464': 4, + '465': 1, + '466': 7, + '467': 14, + '468': 4, + '469': 3, + '470': 2, + '471': 8, + '472': 1, + '473': 1, + '474': 1, + '475': 5, + '476': 3, + '477': 5, + '478': 13, + '479': 4, + '480': 15, + '481': 6, + '482': 20, + '483': 1, + '484': 4, + '485': 1, + '486': 5, + '487': 3, + '488': 1, + '489': 1, + '490': 3, + '491': 1, + '492': 7, + '493': 1, + '494': 5, + '495': 2, + '496': 2, + '497': 2, + '498': 2, + '499': 11, + '500': 1, + '501': 1, + '502': 3, + '503': 7, + '504': 1, + '505': 7, + '506': 2, + '507': 2, + '508': 4, + '509': 2, + '510': 5, + '511': 2, + '513': 5, + '514': 1, + '515': 2, + '516': 2, + '517': 2, + '518': 1, + '519': 1, + '520': 2, + '521': 1, + '522': 1, + '523': 1, + '524': 1, + '525': 3, + '526': 2, + '527': 2, + '528': 2, + '529': 3, + '530': 2, + '531': 1, + '532': 1, + '533': 2, + '534': 2, + '535': 2, + '536': 2, + '537': 2, + '538': 5, + '539': 1, + '540': 5, + '541': 58, + '543': 17, + '544': 22, + '545': 46, + '546': 3, + '547': 2, + '548': 1, + '549': 1, + '550': 2, + '551': 2, + '552': 1, + '553': 2, + '554': 1, + '555': 1, + '556': 4, + '557': 4, + '558': 3, + '559': 2, + '560': 1, + '561': 3, + '562': 1, + '563': 2, + '564': 1, + '565': 4, + '566': 1, + '567': 1, + '568': 2, + '569': 2, + '570': 1, + '571': 1, + '572': 1, + '573': 11, + '574': 1, + '575': 1, + '576': 1, + '577': 1, + '578': 1, + '579': 1, + '580': 1, + '581': 1, + '582': 1, + '583': 1, + '584': 1, + '585': 3, + '586': 1, + '587': 1, + '588': 1, + '589': 1, + '590': 1, + '591': 2, + '592': 3, + '593': 1, + '594': 1, + '595': 1, + '596': 1, + '597': 44, + '598': 1, + '599': 2, + '600': 1, + '601': 10, + '602': 7, + '603': 1, + '604': 7, + '605': 10, + '606': 15, + '607': 1, + '608': 21, + '609': 2, + '610': 10, + '611': 13, + '612': 17, + '613': 1, + '614': 1, + '615': 4, + '616': 43, + '617': 2, + '618': 3, + '619': 14, + '620': 5, + '621': 3, + '622': 6, + '623': 12, + '624': 18, + '625': 5, + '626': 30, + '627': 8, + '628': 14, + '629': 2, + '630': 3, + '631': 2, + '632': 31, + '633': 17, + '634': 21, + '635': 1, + '636': 5, + '638': 17, + '639': 4, + '640': 3, + '641': 12, + '642': 5, + '643': 6, + '644': 6, + '645': 35, + '646': 1, + '647': 2, + '648': 2, + '649': 1, + '650': 1, + '651': 2, + '652': 1, + '653': 1, + '654': 1, + '655': 1, + '656': 3, + '657': 1, + '658': 3, + '659': 3, + '660': 1, + '661': 1, + '662': 2, + '663': 1, + '664': 1, + '665': 1, + '666': 1, + '667': 1, + '668': 1, + '669': 1, + '670': 2, + '671': 1, + '672': 7, + '673': 1, + '674': 1, + '675': 4, + '676': 2, + '677': 1, + '678': 1, + '679': 1, + '680': 2, + '681': 1, + '682': 1, + '683': 2, + '684': 1, + '685': 1, + '686': 2, + '687': 1, + '688': 14, + '689': 4, + '690': 4, + '691': 2, + '692': 2, + '693': 7, + '694': 1, + '695': 1, + '696': 1, + '697': 1, + '698': 1, + '699': 2, + '700': 2, + '701': 1, + '702': 1, + '703': 1, + '704': 2, + '705': 2, + '706': 1, + '707': 1, + '708': 1, + '709': 1, + '710': 1, + '711': 2, + '712': 2, + '713': 1, + '714': 2, + '715': 1, + '716': 1, + '717': 1, + '718': 1, + '719': 1, + '720': 1, + '721': 1, + '722': 1, + '723': 2, + '724': 1, + '725': 1, + '726': 4, + '727': 1, + '728': 1, + '729': 1, + '730': 1, + '731': 1, + '732': 1, + '733': 2, + '734': 1, + '735': 1, + '736': 2, + '737': 2, + '738': 5, + '739': 2, + '740': 1, + '741': 1, + '742': 1, + '743': 1, + '744': 1, + '745': 2, + '746': 2, + '747': 1, + '748': 1, + '749': 1, + '750': 1, + '751': 2, + '752': 1, + '753': 1, + '754': 1, + '755': 2, + '756': 3, + '757': 1, + '758': 1, + '759': 1, + '760': 2, + '761': 1, + '762': 3, + '763': 1, + '764': 4, + '765': 5, + '766': 2, + '767': 2, + '768': 1, + '769': 2, + '770': 3, + '771': 1, + '772': 2, + '773': 1, + '774': 2, + '775': 1, + '776': 1, + '777': 7, + '778': 2, + '779': 1, + '780': 1, + '781': 1, + '782': 1, + '783': 1, + '784': 1, + '785': 1, + '786': 2, + '787': 2, + '788': 1, + '789': 1, + '790': 2, + '791': 1, + '792': 10, + '793': 2, + '794': 2, + '795': 1, + '796': 3, + '797': 5, + '798': 1, + '799': 2, + '800': 1, + '801': 1, + '802': 1, + '803': 1, + '804': 5, + '805': 10, + '806': 1, + '807': 7, + '808': 1, + '809': 1, + '810': 1, + '811': 4, + '812': 5, + '813': 2, + '814': 9, + '815': 1, + '816': 1, + '817': 8, + '818': 2, + '819': 1, + '820': 5, + '821': 1, + '822': 2, + '823': 3, + '824': 11, + '825': 2, + '826': 34, + '827': 1, + '828': 3, + '829': 1, + '830': 1, + '831': 1, + '832': 2, + '833': 2, + '834': 2, + '835': 1, + '836': 1, + '837': 1, + '838': 1, + '839': 20, + '840': 1, + '841': 3, + '842': 16, + '843': 2, + '844': 3, + '845': 1, + '846': 1, + '847': 1, + '848': 1, + '849': 1, + '850': 1, + }, + source_domain: { + 'theguardian.com': 148, + 'nytimes.com': 124, + 'theverge.com': 103, + 'washingtonpost.com': 94, + 'wired.com': 66, + 'bbc.com': 58, + 'reuters.com': 54, + 'arstechnica.com': 52, + 'futurism.com': 52, + 'vice.com': 52, + 'businessinsider.com': 46, + 'forbes.com': 40, + 'apnews.com': 39, + 'dailymail.co.uk': 39, + 'gizmodo.com': 36, + 'wsj.com': 36, + 'telegraph.co.uk': 33, + 'cnn.com': 32, + 'cbsnews.com': 31, + 'independent.co.uk': 31, + 'mashable.com': 30, + 'nbcnews.com': 30, + 'twitter.com': 30, + 'theregister.com': 29, + 'fortune.com': 28, + 'nypost.com': 28, + 'qz.com': 28, + 'engadget.com': 26, + 'abc.net.au': 25, + 'technologyreview.com': 25, + 'cnbc.com': 24, + 'techcrunch.com': 23, + 'usatoday.com': 22, + 'cnet.com': 21, + 'medium.com': 19, + 'npr.org': 19, + 'theregister.co.uk': 15, + 'arxiv.org': 14, + 'bloomberg.com': 14, + 'buzzfeednews.com': 14, + 'foxnews.com': 14, + 'thetimes.co.uk': 14, + 'time.com': 14, + 'digitaltrends.com': 13, + 'dailydot.com': 12, + 'edition.cnn.com': 12, + 'electrek.co': 12, + 'theatlantic.com': 12, + 'thehill.com': 12, + 'venturebeat.com': 12, + 'abcnews.go.com': 11, + 'businessinsider.com.au': 11, + 'latimes.com': 11, + 'news.sky.com': 11, + 'slate.com': 11, + 'thenextweb.com': 11, + 'yahoo.com': 11, + 'gizmodo.com.au': 10, + 'motherboard.vice.com': 10, + 'pcmag.com': 10, + 'propublica.org': 10, + 'techtimes.com': 10, + 'thesun.co.uk': 10, + 'boingboing.net': 9, + 'foxbusiness.com': 9, + 'indiatoday.in': 9, + 'interestingengineering.com': 9, + 'inverse.com': 9, + 'theconversation.com': 9, + 'biometricupdate.com': 8, + 'cbc.ca': 8, + 'extremetech.com': 8, + 'fastcompany.com': 8, + 'smh.com.au': 8, + 'techdirt.com': 8, + 'vox.com': 8, + 'zdnet.com': 8, + 'euronews.com': 7, + 'freep.com': 7, + 'hindustantimes.com': 7, + 'huffingtonpost.com': 7, + 'iflscience.com': 7, + 'inc.com': 7, + 'money.cnn.com': 7, + 'newsweek.com': 7, + 'nymag.com': 7, + 'petapixel.com': 7, + 'siliconangle.com': 7, + 'splinternews.com': 7, + 'stuff.co.nz': 7, + 'thedailybeast.com': 7, + 'timesnownews.com': 7, + 'wionews.com': 7, + 'aclu.org': 6, + 'algorithmwatch.org': 6, + 'aljazeera.com': 6, + 'analyticsindiamag.com': 6, + 'hrw.org': 6, + 'huffpost.com': 6, + 'mirror.co.uk': 6, + 'news.com.au': 6, + 'newser.com': 6, + 'newsnationnow.com': 6, + 'politico.eu': 6, + 'popularmechanics.com': 6, + 'statnews.com': 6, + 'themarkup.org': 6, + 'timesofindia.indiatimes.com': 6, + 'wired.co.uk': 6, + 'axios.com': 5, + 'bbc.co.uk': 5, + 'carscoops.com': 5, + 'cointelegraph.com': 5, + 'cyberscoop.com': 5, + 'economist.com': 5, + 'en.wikipedia.org': 5, + 'entrepreneur.com': 5, + 'france24.com': 5, + 'ft.com': 5, + 'ftc.gov': 5, + 'jalopnik.com': 5, + 'koreaherald.com': 5, + 'metro.co.uk': 5, + 'ndtv.com': 5, + 'newsguardtech.com': 5, + 'papers.ssrn.com': 5, + 'readwrite.com': 5, + 'scmp.com': 5, + 'snopes.com': 5, + 'thedrive.com': 5, + 'theinquirer.net': 5, + 'theintercept.com': 5, + 'torquenews.com': 5, + 'tribune.com.pk': 5, + '404media.co': 4, + 'autoblog.com': 4, + 'autoevolution.com': 4, + 'autonews.com': 4, + 'baijiahao.baidu.com': 4, + 'complex.com': 4, + 'cybernews.com': 4, + 'digitaljournal.com': 4, + 'factcheck.afp.com': 4, + 'foxglove.org.uk': 4, + 'globalnews.ca': 4, + 'huffingtonpost.co.uk': 4, + 'huffingtonpost.com.au': 4, + 'indiatimes.com': 4, + 'insideevs.com': 4, + 'justice.gov': 4, + 'koreatimes.co.kr': 4, + 'kotaku.com': 4, + 'malwaretips.com': 4, + 'mercurynews.com': 4, + 'morningbrew.com': 4, + 'msn.com': 4, + 'nydailynews.com': 4, + 'nzherald.co.nz': 4, + 'patch.com': 4, + 'phys.org': 4, + 'privacyinternational.org': 4, + 'projects.tampabay.com': 4, + 'rt.com': 4, + 'sbs.com.au': 4, + 'spectrum.ieee.org': 4, + 'standard.co.uk': 4, + 'static.nhtsa.gov': 4, + 'straitstimes.com': 4, + 'tapinto.net': 4, + 'tech.co': 4, + 'techrepublic.com': 4, + 'themessenger.com': 4, + 'variety.com': 4, + 'windowscentral.com': 4, + 'youtube.com': 4, + '9to5mac.com': 3, + 'about.fb.com': 3, + 'al-monitor.com': 3, + 'alphr.com': 3, + 'billboard.com': 3, + 'bizjournals.com': 3, + 'blackenterprise.com': 3, + 'bostonglobe.com': 3, + 'businessinsider.in': 3, + 'caranddriver.com': 3, + 'chicago.suntimes.com': 3, + 'choice.com.au': 3, + 'courthousenews.com': 3, + 'dailycaller.com': 3, + 'dailyo.in': 3, + 'democratandchronicle.com': 3, + 'dexerto.com': 3, + 'dmv.ca.gov': 3, + 'eff.org': 3, + 'en.yna.co.kr': 3, + 'fiercehealthcare.com': 3, + 'finance.yahoo.com': 3, + 'globalwitness.org': 3, + 'haaretz.com': 3, + 'hothardware.com': 3, + 'ibtimes.co.uk': 3, + 'infoworld.com': 3, + 'insidehighered.com': 3, + 'insider.com': 3, + 'insurancejournal.com': 3, + 'itechpost.com': 3, + 'itnews.com.au': 3, + 'jezebel.com': 3, + 'lawandcrime.com': 3, + 'lifehacker.com': 3, + 'linkedin.com': 3, + 'livemint.com': 3, + 'logicallyfacts.com': 3, + 'marketwatch.com': 3, + 'mic.com': 3, + 'mindmatters.ai': 3, + 'nature.com': 3, + 'news.bloomberglaw.com': 3, + 'news.ycombinator.com': 3, + 'news18.com': 3, + 'newsbytesapp.com': 3, + 'newscientist.com': 3, + 'newshub.co.nz': 3, + 'nj.com': 3, + 'pandaily.com': 3, + 'people.com': 3, + 'politico.com': 3, + 'polygon.com': 3, + 'protocol.com': 3, + 'reddit.com': 3, + 'restofworld.org': 3, + 'roboticsbusinessreview.com': 3, + 'rollingstone.com': 3, + 'sciencealert.com': 3, + 'sfexaminer.com': 3, + 'sfgate.com': 3, + 'siliconrepublic.com': 3, + 'sixthtone.com': 3, + 'tampabay.com': 3, + 'tech.slashdot.org': 3, + 'techinasia.com': 3, + 'techpolicy.press': 3, + 'the-independent.com': 3, + 'the-sun.com': 3, + 'theautopian.com': 3, + 'thedrum.com': 3, + 'thehindu.com': 3, + 'thestreet.com': 3, + 'thewrap.com': 3, + 'today.com': 3, + 'trustedreviews.com': 3, + 'voanews.com': 3, + 'washingtontimes.com': 3, + 'weforum.org': 3, + 'abajournal.com': 2, + 'abovethelaw.com': 2, + 'acsh.org': 2, + 'adcu.org.uk': 2, + 'advox.globalvoices.org': 2, + 'afr.com': 2, + 'ajc.com': 2, + 'allure.com': 2, + 'amnesty.org': 2, + 'androidauthority.com': 2, + 'aol.com': 2, + 'appleinsider.com': 2, + 'article.wn.com': 2, + 'au.pcmag.com': 2, + 'audacy.com': 2, + 'balkaninsight.com': 2, + 'blog.google': 2, + 'blogs.wsj.com': 2, + 'brookings.edu': 2, + 'brusselstimes.com': 2, + 'businesstoday.in': 2, + 'bustle.com': 2, + 'canadianlawyermag.com': 2, + 'ccn.com': 2, + 'cfodive.com': 2, + 'chicagomag.com': 2, + 'cityandstateny.com': 2, + 'coindesk.com': 2, + 'computerworld.com.au': 2, + 'consumer.org.nz': 2, + 'consumerist.com': 2, + 'cosmopolitan.com': 2, + 'cultofmac.com': 2, + 'cyberdaily.au': 2, + 'dailynews.com': 2, + 'dailywire.com': 2, + 'darkreading.com': 2, + 'dataconomy.com': 2, + 'decrypt.co': 2, + 'denver7.com': 2, + 'dezeen.com': 2, + 'digitalethics.org': 2, + 'digitalspy.com': 2, + 'disabilitynewsservice.com': 2, + 'driveteslacanada.ca': 2, + 'driving.co.uk': 2, + 'dukechronicle.com': 2, + 'edmontonpolice.ca': 2, + 'ekker.legal': 2, + 'euractiv.com': 2, + 'express.co.uk': 2, + 'facebook.com': 2, + 'financialexpress.com': 2, + 'floridapolitics.com': 2, + 'fox4news.com': 2, + 'frontofficesports.com': 2, + 'fullfact.org': 2, + 'gearbrain.com': 2, + 'geek.com': 2, + 'geekwire.com': 2, + 'giantfreakinrobot.com': 2, + 'github.com': 2, + 'glamour.com': 2, + 'globalvillagespace.com': 2, + 'govtech.com': 2, + 'graphika.com': 2, + 'hackingdistributed.com': 2, + 'hackread.com': 2, + 'hai.stanford.edu': 2, + 'hani.co.kr': 2, + 'hypebeast.com': 2, + 'ic3.gov': 2, + 'independent.ie': 2, + 'inews.co.uk': 2, + 'infowars.com': 2, + 'inputmag.com': 2, + 'inquisitr.com': 2, + 'intellinews.com': 2, + 'internetofbusiness.com': 2, + 'iol.co.za': 2, + 'irishtimes.com': 2, + 'isdglobal.org': 2, + 'jpost.com': 2, + 'kffhealthnews.org': 2, + 'kiro7.com': 2, + 'knightlawgroup.com': 2, + 'ktla.com': 2, + 'lalibre.be': 2, + 'latestly.com': 2, + 'lexology.com': 2, + 'link.springer.com': 2, + 'lsj.com.au': 2, + 'mediamatters.org': 2, + 'min.news': 2, + 'motherjones.com': 2, + 'nationalpost.com': 2, + 'nationalreview.com': 2, + 'nbcbayarea.com': 2, + 'news.abplive.com': 2, + 'news.trust.org': 2, + 'newsone.com': 2, + 'newyorker.com': 2, + 'nme.com': 2, + 'nola.com': 2, + 'notebookcheck.net': 2, + 'ntsb.gov': 2, + 'openai.com': 2, + 'pagesix.com': 2, + 'pcworld.com': 2, + 'pnas.org': 2, + 'politifact.com': 2, + 'popsci.com': 2, + 'post-gazette.com': 2, + 'pymnts.com': 2, + 'rappler.com': 2, + 'reason.com': 2, + 'recode.net': 2, + 'researchgate.net': 2, + 'revealnews.org': 2, + 'sacbee.com': 2, + 'screenshot-media.com': 2, + 'sea.mashable.com': 2, + 'seanbmcgregor.com': 2, + 'searchengineland.com': 2, + 'seattletimes.com': 2, + 'sec.gov': 2, + 'seekingalpha.com': 2, + 'sfchronicle.com': 2, + 'sfist.com': 2, + 'simonwillison.net': 2, + 'sitepronews.com': 2, + 'spiegel.de': 2, + 'startribune.com': 2, + 'taiwannews.com.tw': 2, + 'techmonitor.ai': 2, + 'techradar.com': 2, + 'techspot.com': 2, + 'techtransparencyproject.org': 2, + 'techxplore.com': 2, + 'the-decoder.com': 2, + 'thebulletin.org': 2, + 'thecut.com': 2, + 'thediplomat.com': 2, + 'theinformation.com': 2, + 'thenewstack.io': 2, + 'thequint.com': 2, + 'thestar.com': 2, + 'tiktok.com': 2, + 'tomsguide.com': 2, + 'turkishminute.com': 2, + 'uk.news.yahoo.com': 2, + 'uni-watch.com': 2, + 'unilad.com': 2, + 'usnews.com': 2, + 'vanderbilthustler.com': 2, + 'vietnamnet.vn': 2, + 'voicebot.ai': 2, + 'walesonline.co.uk': 2, + 'wral.com': 2, + 'youtu.be': 2, + '10news.com': 1, + '11alive.com': 1, + '2025ad.com': 1, + '71republic.com': 1, + '947.co.za': 1, + '972mag.com': 1, + '9news.com': 1, + '9news.com.au': 1, + '9to5google.com': 1, + 'aa.com.tr': 1, + 'aaai.org': 1, + 'aap.com.au': 1, + 'abc15.com': 1, + 'abc4.com': 1, + 'abc7.com': 1, + 'abc7news.com': 1, + 'abc7ny.com': 1, + 'abplive.in': 1, + 'accessnow.org': 1, + 'aclunc.org': 1, + 'advrider.com': 1, + 'adweek.com': 1, + 'aebsettlement.com': 1, + 'afrotech.com': 1, + 'agora.md': 1, + 'aibusiness.com': 1, + 'aisnakeoil.substack.com': 1, + 'alaskasnewssource.com': 1, + 'allaboutai.com': 1, + 'amp-theguardian-com.cdn.ampproject.org': 1, + 'analyticsdrift.com': 1, + 'analyticsvidhya.com': 1, + 'androidheadlines.com': 1, + 'antoinespeaks.co.uk': 1, + 'aplus.com': 1, + 'app.com': 1, + 'appleworld.today': 1, + 'archdaily.com': 1, + 'archive.boston.com': 1, + 'artnews.com': 1, + 'asiafinancial.com': 1, + 'atlanticcouncil.org': 1, + 'au.finance.yahoo.com': 1, + 'augustman.com': 1, + 'autoguide.com': 1, + 'autonews.gasgoo.com': 1, + 'autosafety.org': 1, + 'awfulannouncing.com': 1, + 'azcentral.com': 1, + 'azfamily.com': 1, + 'bair.berkeley.edu': 1, + 'bangkokpost.com': 1, + 'bankinfosecurity.com': 1, + 'barrysookman.com': 1, + 'barstoolsports.com': 1, + 'bartlesvilleradio.com': 1, + 'beckersasc.com': 1, + 'beckershospitalreview.com': 1, + 'beincrypto.com': 1, + 'belairassociation.org': 1, + 'belganewsagency.eu': 1, + 'benitolink.com': 1, + 'benzinga.com': 1, + 'biggerlawfirm.com': 1, + 'bipartisanpolicy.org': 1, + 'bitdefender.com': 1, + 'bits.blogs.nytimes.com': 1, + 'blavity.com': 1, + 'blitzquotidiano.it': 1, + 'blockchain-council.org': 1, + 'blockchaintechnology-news.com': 1, + 'blog.angryasianman.com': 1, + 'blog.bity.com': 1, + 'blog.botego.com': 1, + 'blog.coinbase.com': 1, + 'blog.conceptnet.io': 1, + 'blog.galalaw.com': 1, + 'blog.openai.com': 1, + 'blog.slock.it': 1, + 'blog.twitter.com': 1, + 'blogs.bing.com': 1, + 'blogs.discovermagazine.com': 1, + 'blogs.microsoft.com': 1, + 'bloomberg.com.': 1, + 'bnnbloomberg.ca': 1, + 'bnnbreaking.com': 1, + 'boredpanda.com': 1, + 'boston.com': 1, + 'bostonherald.com': 1, + 'boxmining.com': 1, + 'br.de': 1, + 'brecorder.com': 1, + 'breitbart.com': 1, + 'bronx.news12.com': 1, + 'business-standard.com': 1, + 'businesscloud.co.uk': 1, + 'businessnewsdaily.com': 1, + 'businesstimes.com.sg': 1, + 'businesswire.com': 1, + 'bussgeldportal.de': 1, + 'buzzfeed.com': 1, + 'ca.finance.yahoo.com': 1, + 'ca.movies.yahoo.com': 1, + 'ca.news.yahoo.com': 1, + 'cacm.acm.org': 1, + 'cafemom.com': 1, + 'caixinglobal.com': 1, + 'capitalbrief.com': 1, + 'capradio.org': 1, + 'carcomplaints.com': 1, + 'carloscreusmoreira.medium.com': 1, + 'carthrottle.com': 1, + 'casetext.com': 1, + 'catonetworks.com': 1, + 'cbs12.com': 1, + 'cbs17.com': 1, + 'ccjdigital.com': 1, + 'cdn.arstechnica.net': 1, + 'cdn.openai.com': 1, + 'cdn.sanity.io': 1, + 'cdotrends.com': 1, + 'cdt.org': 1, + 'cepa.org': 1, + 'cfr.org': 1, + 'cgdev.org': 1, + 'change.org': 1, + 'channel4.com': 1, + 'channels.theinnovationenterprise.com': 1, + 'chatbotslife.com': 1, + 'chatbotsmagazine.com': 1, + 'chicagoreader.com': 1, + 'chicagotribune.com': 1, + 'chinadaily.com.cn': 1, + 'christiantoday.com': 1, + 'cinemablend.com': 1, + 'cio.com': 1, + 'cityam.com': 1, + 'citylab.com': 1, + 'civilbeat.org': 1, + 'cjr.org': 1, + 'clarin.com': 1, + 'clarksonlawfirm.com': 1, + 'claytonutz.com': 1, + 'click2houston.com': 1, + 'cmu.edu': 1, + 'codastory.com': 1, + 'cohenmilstein.com': 1, + 'coincentral.com': 1, + 'coincodex.com': 1, + 'coinmarketcap.com': 1, + 'colombiaone.com': 1, + 'coloradohometownweekly.com': 1, + 'colorlines.com': 1, + 'commondreams.org': 1, + 'comptroller.nyc.gov': 1, + 'computerworld.com': 1, + 'computing.co.uk': 1, + 'consumersinternational.org': 1, + 'context-cdn.washingtonpost.com': 1, + 'copyrightlately.com': 1, + 'corporatefinanceinstitute.com': 1, + 'cosmeticsbusiness.com': 1, + 'cosmosmagazine.com': 1, + 'coyotecountrylv.com': 1, + 'cpomagazine.com': 1, + 'cpsc.gov': 1, + 'cracked.com': 1, + 'creativebloq.com': 1, + 'cryptocompare.com': 1, + 'cryptocurrencyhub.io': 1, + 'cryptopolitan.com': 1, + 'csmapnyu.org': 1, + 'csoonline.com': 1, + 'ctinsider.com': 1, + 'ctvnews.ca': 1, + 'culvercityobserver.com': 1, + 'curiosity.com': 1, + 'cyber.fsi.stanford.edu': 1, + 'cybersecurityfordemocracy.org': 1, + 'd.dailynews.co.th': 1, + 'dailybreeze.com': 1, + 'dailycitizen.focusonthefamily.com': 1, + 'dailypress.net': 1, + 'dailystar.co.uk': 1, + 'dallasnews.com': 1, + 'danrl.com': 1, + 'dataprivacylab.org': 1, + 'deadline.com': 1, + 'dealstreetasia.com': 1, + 'dearauthor.com': 1, + 'deccanherald.com': 1, + 'delish.com': 1, + 'denverpost.com': 1, + 'designnews.com': 1, + 'detroitnews.com': 1, + 'digg.com': 1, + 'digit.fyi': 1, + 'digitalcameraworld.com': 1, + 'dispatch.com': 1, + 'disputeresolutiongermany.com': 1, + 'dkb.blog': 1, + 'dl.acm.org': 1, + 'dmlp.org': 1, + 'dmv.com': 1, + 'dnaindia.com': 1, + 'docs.rwu.edu': 1, + 'donga.com': 1, + 'dotesports.com': 1, + 'dpreview.com': 1, + 'dttc.sggp.org.vn': 1, + 'dw.com': 1, + 'ebaumsworld.com': 1, + 'ecns.cn': 1, + 'economictimes.indiatimes.com': 1, + 'econotimes.com': 1, + 'editorialge.com': 1, + 'edpb.europa.eu': 1, + 'edrants.com': 1, + 'eldiario.es': 1, + 'electronicproducts.com': 1, + 'elitereaders.com': 1, + 'elle.com': 1, + 'elmundo.es': 1, + 'elobservador.com.uy': 1, + 'emirates247.com': 1, + 'en.dailypakistan.com.pk': 1, + 'en.econostrum.info': 1, + 'en.wikinews.org': 1, + 'energyinfrapost.com': 1, + 'english.elpais.com': 1, + 'english.jagran.com': 1, + 'ent.siteintelgroup.com': 1, + 'epic.org': 1, + 'escapistmagazine.com': 1, + 'eteknix.com': 1, + 'etftrends.com': 1, + 'eticasfoundation.org': 1, + 'eu.usatoday.com': 1, + 'eulawenforcement.com': 1, + 'eurogamer.net': 1, + 'europepmc.org': 1, + 'explica.co': 1, + 'expressnews.com': 1, + 'fairfieldsuntimes.com': 1, + 'fairworkweek.org': 1, + 'familyhandyman.com': 1, + 'features.propublica.org': 1, + 'finance.nine.com.au': 1, + 'financialregnews.com': 1, + 'finbold.com': 1, + 'firstpost.com': 1, + 'flawedfacedata.com': 1, + 'fool.com': 1, + 'forbesafrica.com': 1, + 'fordfoundation.org': 1, + 'foreignpolicy.com': 1, + 'forum.facepunch.com': 1, + 'forward.com': 1, + 'fossa.com': 1, + 'foundation.mozilla.org': 1, + 'fox2detroit.com': 1, + 'fox5dc.com': 1, + 'fox5ny.com': 1, + 'fox5sandiego.com': 1, + 'foxcarolina.com': 1, + 'freakonomics.com': 1, + 'freemovement.org.uk': 1, + 'fudzilla.com': 1, + 'fullerproject.org': 1, + 'futureoflife.org': 1, + 'futureparty.com': 1, + 'gamasutra.com': 1, + 'gamedaily.biz': 1, + 'gamespot.com': 1, + 'gameworldobserver.com': 1, + 'garyrubinstein.teachforus.org': 1, + 'gawker.com': 1, + 'geekologie.com': 1, + 'gffbrokers.com': 1, + 'gineersnow.com': 1, + 'github.blog': 1, + 'githubcopilotlitigation.com': 1, + 'gizchina.com': 1, + 'gizmodo.co.uk': 1, + 'glamourmagazine.co.uk': 1, + 'globalcitizen.org': 1, + 'globalnation.inquirer.net': 1, + 'globaltimes.cn': 1, + 'globalvoices.org': 1, + 'gmauthority.com': 1, + 'gmw.cn': 1, + 'gofundme.com': 1, + 'goodmorningamerica.com': 1, + 'gothamist.com': 1, + 'gov.uk': 1, + 'gq.com': 1, + 'graphics.wsj.com': 1, + 'graziadaily.co.uk': 1, + 'greekreporter.com': 1, + 'greenbot.com': 1, + 'gwern.net': 1, + 'gzeromedia.com': 1, + 'hackaday.com': 1, + 'hackernoon.com': 1, + 'hardworkingtrucks.com': 1, + 'hawaiinewsnow.com': 1, + 'hawaiitribune-herald.com': 1, + 'hbr.org': 1, + 'hbswk.hbs.edu': 1, + 'hcamag.com': 1, + 'healthcarefinancenews.com': 1, + 'heise.de': 1, + 'helpnetsecurity.com': 1, + 'heraldnews.com': 1, + 'hindi.news18.com': 1, + 'hoit.uk': 1, + 'home.bt.com': 1, + 'hotelmanagement.net': 1, + 'howtogeek.com': 1, + 'hrmonline.com.au': 1, + 'hrtechnologynews.com': 1, + 'hstoday.us': 1, + 'huffingtonpost.ca': 1, + 'huntonprivacyblog.com': 1, + 'hurriyetdailynews.com': 1, + 'hypebae.com': 1, + 'ia.acs.org.au': 1, + 'iafrica.com': 1, + 'iafrikan.com': 1, + 'ibtimes.com': 1, + 'idiallo.com': 1, + 'ilfattoquotidiano.it': 1, + 'ilpost.it': 1, + 'imd.org': 1, + 'imore.com': 1, + 'in.finance.yahoo.com': 1, + 'in.mashable.com': 1, + 'incidentdatabase.ai': 1, + 'india.com': 1, + 'indianexpress.com': 1, + 'indiatvnews.com': 1, + 'industryweek.com': 1, + 'indystar.com': 1, + 'information-age.com': 1, + 'innotechtoday.com': 1, + 'inshorts.com': 1, + 'insideedition.com': 1, + 'insidehook.com': 1, + 'insideprivacy.com': 1, + 'insights.globalspec.com': 1, + 'insights.tmpw.co.uk': 1, + 'instagram.com': 1, + 'intelligenttransport.com': 1, + 'intouchweekly.com': 1, + 'investors.com': 1, + 'io9.gizmodo.com': 1, + 'iottechnews.com': 1, + 'iotworldtoday.com': 1, + 'iphonehacks.com': 1, + 'ipi.media': 1, + 'ipn.md': 1, + 'ipvm.com': 1, + 'irishmirror.ie': 1, + 'ishn.com': 1, + 'itwire.com': 1, + 'ivebeenmugged.typepad.com': 1, + 'jcwi.org.uk': 1, + 'jdsupra.com': 1, + 'jsjc.gov.cn': 1, + 'jta.org': 1, + 'judiciary.uk': 1, + 'justcareusa.org': 1, + 'justjared.com': 1, + 'justsomething.co': 1, + 'k.sina.cn': 1, + 'kathmandupost.ekantipur.com': 1, + 'kfm.co.za': 1, + 'khaleejtimes.com': 1, + 'khq.com': 1, + 'kingsleynapley.co.uk': 1, + 'knowyourmeme.com': 1, + 'kotaku.co.uk': 1, + 'kr-asia.com': 1, + 'krdo.com': 1, + 'krem.com': 1, + 'kslnewsradio.com': 1, + 'ktvu.com': 1, + 'kurdistan24.net': 1, + 'labsnews.com': 1, + 'ladbible.com': 1, + 'lagrandeobserver.com': 1, + 'laion.ai': 1, + 'lamag.com': 1, + 'lasvegassun.com': 1, + 'latimesblogs.latimes.com': 1, + 'latintimes.com': 1, + 'laweekly.com': 1, + 'lawenforcementtoday.com': 1, + 'lawfaremedia.org': 1, + 'lawgazette.co.uk': 1, + 'learningenglish.voanews.com': 1, + 'legalreader.com': 1, + 'lemonde.fr': 1, + 'lentrepreneur.co': 1, + 'lesswrong.com': 1, + 'lgbtqnation.com': 1, + 'liaa.dc.uba.ar': 1, + 'lifesitenews.com': 1, + 'livescience.com': 1, + 'lobste.rs': 1, + 'local12.com': 1, + 'losangelesblade.com': 1, + 'lr21.com.uy': 1, + 'macarthurjustice.org': 1, + 'macrumors.com': 1, + 'madisonrecord.com': 1, + 'makeuseof.com': 1, + 'malaymail.com': 1, + 'malicious.life': 1, + 'mansworldindia.com': 1, + 'marginalrevolution.com': 1, + 'marketplace.org': 1, + 'markey.senate.gov': 1, + 'martinmontilino.com': 1, + 'massdevice.com': 1, + 'massivelyop.com': 1, + 'mcknightshomecare.com': 1, + 'mcvuk.com': 1, + 'me.pcmag.com': 1, + 'mediaite.com': 1, + 'medianama.com': 1, + 'medicaleconomics.com': 1, + 'medium.freecodecamp.org': 1, + 'medscape.com': 1, + 'mentalfloss.com': 1, + 'meta-writer.livejournal.com': 1, + 'meta.stackoverflow.com': 1, + 'metalinjection.net': 1, + 'metrorailnews.in': 1, + 'metroweekly.com': 1, + 'meyerweb.com': 1, + 'miamiherald.com': 1, + 'michaeleisen.org': 1, + 'michiganradio.org': 1, + 'middleeasteye.net': 1, + 'military.com': 1, + 'mlive.com': 1, + 'mobility21.cmu.edu': 1, + 'mondaq.com': 1, + 'montevideo.com.uy': 1, + 'montrealgazette.com': 1, + 'mothership.sg': 1, + 'motorbiscuit.com': 1, + 'motormoutharabia.com': 1, + 'motortrend.com': 1, + 'msnbc.com': 1, + 'multichain.com': 1, + 'musically.com': 1, + 'muycomputerpro.com': 1, + 'myfox28columbus.com': 1, + 'mymodernmet.com': 1, + 'mynorthwest.com': 1, + 'nafcu.org': 1, + 'nasdaq.com': 1, + 'natbuckley.co.uk': 1, + 'nation.com.pk': 1, + 'natlawreview.com': 1, + 'nautil.us': 1, + 'nbc15.com': 1, + 'nbcboston.com': 1, + 'nbcchicago.com': 1, + 'nbcdfw.com': 1, + 'nbclosangeles.com': 1, + 'nbcmiami.com': 1, + 'nbcnewyork.com': 1, + 'nbcphiladelphia.com': 1, + 'nbcwashington.com': 1, + 'nec.com': 1, + 'neil-clarke.com': 1, + 'neoskosmos.com': 1, + 'neowin.net': 1, + 'newindianexpress.com': 1, + 'newjerseymonitor.com': 1, + 'newmatilda.com': 1, + 'newrepublic.com': 1, + 'news.avclub.com': 1, + 'news.cgtn.com': 1, + 'news.cornell.edu': 1, + 'news.trendmicro.com': 1, + 'news.tvbs.com.tw': 1, + 'news.yahoo.com': 1, + 'news5cleveland.com': 1, + 'news9live.com': 1, + 'newsbeezer.com': 1, + 'newsbtc.com': 1, + 'newsfolo.com': 1, + 'newsmax.com': 1, + 'newsobserver.com': 1, + 'newson6.com': 1, + 'newsroom.gettyimages.com': 1, + 'newstatesman.com': 1, + 'newswise.com': 1, + 'newsx.com': 1, + 'nguoi-viet.com': 1, + 'nickdiakopoulos.com': 1, + 'nj1015.com': 1, + 'nknews.org': 1, + 'northjersey.com': 1, + 'nouse.co.uk': 1, + 'nowthisnews.com': 1, + 'nst.com.my': 1, + 'ntd.com': 1, + 'ny1.com': 1, + 'observers.france24.com': 1, + 'ocregister.com': 1, + 'offthemainpage.com': 1, + 'olhardigital.com.br': 1, + 'opb.org': 1, + 'osf.io': 1, + 'out-law.com': 1, + 'overlayfalseclaims.com': 1, + 'ovic.vic.gov.au': 1, + 'palestinechronicle.com': 1, + 'pantip.com': 1, + 'parentadvocates.org': 1, + 'pastemagazine.com': 1, + 'pbs.org': 1, + 'pcgamer.com': 1, + 'pewtrusts.org': 1, + 'phonearena.com': 1, + 'physicstoday.scitation.org': 1, + 'plagiarismtoday.com': 1, + 'plantservices.com': 1, + 'pocket-lint.com': 1, + 'policyoptions.irpp.org': 1, + 'poliisi.fi': 1, + 'polizei.bayern.de': 1, + 'polyesterstudio.com': 1, + 'populardemocracy.org': 1, + 'postcourier.com.pg': 1, + 'poynter.org': 1, + 'pressgazette.co.uk': 1, + 'pride.com': 1, + 'princeton.edu': 1, + 'priv.gc.ca': 1, + 'prnewswire.com': 1, + 'proactiveinvestors.com': 1, + 'probonoaustralia.com.au': 1, + 'projects.propublica.org': 1, + 'psmag.com': 1, + 'psychiatrist.com': 1, + 'publiclawproject.org.uk': 1, + 'pulitzercenter.org': 1, + 'pulsenews.co.kr': 1, + 'punchbowl.news': 1, + 'punto-informatico.it': 1, + 'qcnews.com': 1, + 'queerinai.com': 1, + 'queerty.com': 1, + 'queue.acm.org': 1, + 'quinyx.com': 1, + 'radionz.co.nz': 1, + 'raillc.substack.com': 1, + 'rand.org': 1, + 'read.dukeupress.edu': 1, + 'readlion.com': 1, + 'recordedfuture.com': 1, + 'republicworld.com': 1, + 'research.checkpoint.com': 1, + 'resemble.ai': 1, + 'restaurantbusinessonline.com': 1, + 'retailwire.com': 1, + 'retractionwatch.com': 1, + 'rsk.co': 1, + 'rss.onlinelibrary.wiley.com': 1, + 'rts.ch': 1, + 'rtve.es': 1, + 'russiamatters.org': 1, + 's.weibo.com': 1, + 'safiyaunoble.files.wordpress.com': 1, + 'salon.com': 1, + 'samaa.tv': 1, + 'sbnation.com': 1, + 'schneier.com': 1, + 'science.org': 1, + }, + 'CSETv1.AI Task': { + 'facial recognition': 131, + navigation: 112, + patrolling: 62, + 'virtual assistant technology': 58, + security: 57, + 'object detection': 55, + 'autonomous driving': 54, + 'self-driving': 53, + 'object recognition': 51, + 'chat bot': 49, + 'text generation': 46, + 'search optimization': 44, + 'autonomous navigation': 43, + 'search suggestion': 41, + 'Image search': 40, + 'predict recidivism': 38, + 'Rank Applicants': 37, + translation: 37, + prediction: 35, + 'self driving': 35, + 'semi-autonomous navigation': 35, + surveillance: 35, + Driving: 34, + 'resume screening': 33, + 'content generation': 31, + 'deepfake video generation': 31, + autocomplete: 30, + 'content moderation': 30, + recommender: 30, + 'content ranking': 29, + 'Generate Captions': 28, + assembly: 28, + 'image identification': 28, + 'image interpretation': 28, + production: 28, + 'personalized online advertising': 27, + 'image classification': 25, + 'jaywalking detection': 25, + Autopilot: 24, + 'Image Tagging': 24, + 'identity verification': 24, + 'image categorization': 24, + 'obstacle avoidance': 24, + 'route optimization': 24, + filter: 23, + 'image modification': 23, + 'photo edit': 23, + 'Generate Replies': 22, + chatbot: 22, + 'face detection': 22, + 'search engine optimization': 19, + 'Predict Crimes': 13, + 'predictive policing': 13, + recommendation: 13, + generation: 12, + 'detect gunshots': 11, + 'identify weapon calibre': 11, + 'locate gunshots': 11, + 'personalized online search results': 11, + 'population of characteristics for NPCs in a video game': 11, + 'predict shooter movement': 11, + 'Image Analysis': 10, + 'predict store traffic': 10, + 'productivity optimization': 10, + scheduling: 10, + 'image generation': 9, + 'toxicity detection': 9, + 'voice recognition': 9, + 'augmented reality (AR) game Pokémon Go': 8, + 'natural language processing': 8, + 'speech recognition': 8, + 'Design Phone Cases': 7, + 'assign risk': 7, + classification: 7, + 'deepfake audio generation': 7, + 'predict healthcare needs': 7, + 'sentiment analysis': 7, + editing: 6, + 'fraud risk prediction': 6, + 'identify hate speech': 6, + 'webpage maintenance': 6, + 'Crop Images': 5, + 'Personal Assistant': 5, + 'automatic emergency braking': 5, + 'content recommendation': 5, + 'image cropping': 5, + 'security monitor': 5, + 'speech interpretation': 5, + 'Digital Assistant': 4, + 'application screening': 4, + 'audit selection': 4, + 'candidate assessment': 4, + 'diagnose sepsis': 4, + 'emotion recognition': 4, + 'estimating house prices': 4, + 'facial expression recognition': 4, + 'home valuation': 4, + 'identify sepsis': 4, + 'object classification': 4, + 'price determination': 4, + 'product promotion': 4, + 'real estate market forecasting': 4, + 'search engine': 4, + 'search result ranking': 4, + 'sepsis prediction': 4, + transportation: 4, + 'virtual interview': 4, + 'worker management': 4, + delivery: 3, + 'ethical decision making': 3, + 'gender classification': 3, + 'move shelves': 3, + 'performance tracking': 3, + 'reply to questions': 3, + 'video recommender': 3, + 'water quality testing': 3, + 'worker monitoring': 3, + 'assess applicants': 2, + 'audio deepfake': 2, + 'automated speech recognition': 2, + 'automatic tracking': 2, + 'ball detection': 2, + 'chant scripture': 2, + cheerleading: 2, + concierge: 2, + 'driver surveillance': 2, + 'emotion detection': 2, + 'entertain guests': 2, + 'event detection': 2, + 'event prediction': 2, + 'exercise demonstration': 2, + 'eye tracking': 2, + 'facial grouping': 2, + 'household administration': 2, + 'identification or detection': 2, + 'product recognition': 2, + 'rank the reliability of workers': 2, + 'risk assessment': 2, + 'safe driving detection': 2, + 'semi-autonomous driving': 2, + 'shift assignment': 2, + 'speech-to-text': 2, + suggestion: 2, + 'weapon detection': 2, + 'welfare determination': 2, + 'ADA website compliance': 1, + 'Burmese to English translation': 1, + Matching: 1, + 'accessibility compliance': 1, + 'advertisement screening': 1, + 'autonomous food delivery': 1, + 'benefits allocation': 1, + 'brand safety detection': 1, + 'content suggestion': 1, + 'customer service': 1, + database: 1, + 'detect policy violations': 1, + 'enforcement of community guidelines': 1, + 'engagement optimization': 1, + 'exam proctoring': 1, + 'fact-check': 1, + 'human facial image transformation': 1, + 'identify misinformation': 1, + 'image editing': 1, + 'image organization': 1, + 'image recognition': 1, + 'image splicing': 1, + 'job screening': 1, + 'law enforcement': 1, + 'license plate matching': 1, + 'license plate recognition': 1, + 'marketing material generation': 1, + 'optimize movement to win a computer game': 1, + 'passport photo quality check': 1, + 'performance prediction': 1, + 'predict grades on exam': 1, + 'predict necessary amount of medical coverage': 1, + 'predict recidivism risk': 1, + 'recipe generation': 1, + 'recruitment screening': 1, + 'remote proctoring': 1, + 'risk prediction': 1, + robotics: 1, + 'spam filter': 1, + 'stoplight recognition': 1, + 'vaccine allocation': 1, + 'video classification': 1, + 'video suggestion': 1, + 'voice alteration': 1, + 'weapons detection': 1, + 'word association': 1, + }, + 'CSETv1.Deployed': { + yes: 1442, + no: 103, + maybe: 37, + }, + 'CSETv1.Embedded': { + no: 976, + yes: 592, + maybe: 14, + }, + 'CSETv1.Injuries': { + '0': 1480, + '1': 66, + '2': 5, + '10': 27, + '40': 1, + '54': 17, + '416': 3, + '1391': 12, + '55000': 3, + }, + classifications: { + 'CSETv1_Annotator-1:User Test in Controlled Conditions:no': 1841, + 'CSETv1_Annotator-1:Impact on Critical Services:no': 1799, + 'CSETv1_Annotator-1:Multiple AI Interaction:no': 1797, + 'CSETv1_Annotator-1:Producer Test in Controlled Conditions:no': 1785, + 'CSETv1_Annotator-1:Report, Test, or Study of data:no': 1755, + 'CSETv1_Annotator-1:Deployed:yes': 1715, + 'CSETv1_Annotator-1:Involving Minor:no': 1701, + 'CSETv1_Annotator-1:Rights Violation:no': 1690, + 'CSETv1_Annotator-1:Producer Test in Operational Conditions:no': 1683, + 'CSETv1_Annotator-1:User Test in Operational Conditions:no': 1567, + 'CSETv1:User Test in Controlled Conditions:no': 1555, + 'CSETv1_Annotator-1:Intentional Harm:No. Not intentionally designed to perform harm': 1542, + 'CSETv1_Annotator-1:Public Sector Deployment:no': 1526, + 'CSETv1:Multiple AI Interaction:no': 1525, + 'CSETv1:Report, Test, or Study of data:no': 1511, + 'CSETv1:Producer Test in Controlled Conditions:no': 1509, + 'CSETv1_Annotator-1:Detrimental Content:no': 1499, + 'CSETv1_Annotator-1:Entertainment Industry:no': 1498, + 'CSETv1_Annotator-1:AI System:yes': 1497, + 'CSETv1:Impact on Critical Services:no': 1474, + 'CSETv1:Involving Minor:no': 1469, + 'CSETv1:Intentional Harm:No. Not intentionally designed to perform harm': 1462, + 'CSETv1:Producer Test in Operational Conditions:no': 1453, + 'CSETv1:Deployed:yes': 1442, + 'CSETv1:Clear link to Technology:yes': 1425, + 'CSETv1_Annotator-1:Clear link to Technology:yes': 1422, + 'CSETv1:Entertainment Industry:no': 1401, + 'CSETv1:User Test in Operational Conditions:no': 1367, + 'CSETv1:Rights Violation:no': 1332, + 'CSETv1:Public Sector Deployment:no': 1324, + 'CSETv1:Detrimental Content:no': 1281, + 'CSETv1_Annotator-1:Protected Characteristic:no': 1270, + 'CSETv1_Annotator-1:Harm Distribution Basis:none': 1223, + 'CSETv1:AI System:yes': 1221, + 'CSETv1_Annotator-1:Embedded:no': 1206, + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:no': 1192, + 'CSETv1_Annotator-1:AI Harm Level:none': 1153, + 'CSETv1_Annotator-1:Physical Objects:no': 1150, + 'CSETv1_Annotator-1:Special Interest Intangible Harm:no': 1087, + 'CSETv1:Protected Characteristic:no': 1070, + 'CSETv1:Harm Distribution Basis:none': 1059, + 'CSETv1:AI Harm Level:none': 1003, + 'CSETv1:Embedded:no': 976, + 'CSETv1:Annotator’s AI special interest intangible harm assessment:no': 949, + 'CSETv1:Physical Objects:no': 946, + 'CSETv1_Annotator-1:Location Country (two letters):US': 946, + 'CSETv0:Intent:Accident': 935, + 'CSETv1_Annotator-2:User Test in Controlled Conditions:no': 930, + 'CSETv0:Relevant AI functions:Cognition': 924, + 'CSETv1_Annotator-2:Entertainment Industry:no': 908, + 'CSETv1_Annotator-2:Multiple AI Interaction:no': 904, + 'CSETv1_Annotator-2:Producer Test in Operational Conditions:no': 904, + 'CSETv1_Annotator-2:Impact on Critical Services:no': 901, + 'CSETv1_Annotator-1:Location Region:North America': 899, + 'CSETv0:Nature of End User:Amateur': 896, + 'CSETv0:Relevant AI functions:Perception': 892, + 'CSETv1_Annotator-2:Producer Test in Controlled Conditions:no': 884, + 'CSETv1_Annotator-2:Report, Test, or Study of data:no': 870, + 'CSETv1:Special Interest Intangible Harm:no': 862, + 'CSETv1_Annotator-1:Autonomy Level:Autonomy1': 852, + 'CSETv1_Annotator-2:Intentional Harm:No. Not intentionally designed to perform harm': 850, + 'CSETv1:Autonomy Level:Autonomy1': 847, + 'CSETv1_Annotator-2:Involving Minor:no': 843, + 'CSETv1_Annotator-2:Deployed:yes': 841, + 'CSETv1_Annotator-2:Rights Violation:no': 826, + 'CSETv1_Annotator-2:Public Sector Deployment:no': 819, + 'CSETv1_Annotator-2:Detrimental Content:no': 813, + 'CSETv1_Annotator-1:Tangible Harm:no tangible harm, near-miss, or issue': 794, + 'CSETv1:Location Region:North America': 781, + 'CSETv1_Annotator-2:User Test in Operational Conditions:no': 773, + 'CSETv1:Location Country (two letters):US': 769, + 'CSETv0:Relevant AI functions:Action': 768, + 'CSETv1_Annotator-1:Special Interest Intangible Harm:yes': 732, + 'CSETv1_Annotator-1:Tangible Harm:tangible harm definitively occurred': 719, + 'CSETv0:Physical System:Software only': 704, + 'CSETv1_Annotator-2:AI Harm Level:none': 704, + 'CSETv1:Special Interest Intangible Harm:yes': 703, + 'CSETv1_Annotator-2:Protected Characteristic:no': 689, + 'CSETv1:Tangible Harm:tangible harm definitively occurred': 678, + 'CSETv1_Annotator-2:AI System:yes': 675, + 'CSETv1_Annotator-1:Physical Objects:yes': 664, + 'CSETv1_Annotator-2:Embedded:no': 649, + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:yes': 647, + 'CSETv1_Annotator-1:Embedded:yes': 643, + 'CSETv1_Annotator-1:Sector of Deployment:information and communication': 639, + 'CSETv0:Severity:Negligible': 634, + 'CSETv1:Tangible Harm:no tangible harm, near-miss, or issue': 618, + 'CSETv1:Physical Objects:yes': 615, + 'CSETv1:Annotator’s AI special interest intangible harm assessment:yes': 609, + 'CSETv1_Annotator-2:Annotator’s AI special interest intangible harm assessment:no': 607, + 'CSETv1_Annotator-2:Harm Distribution Basis:none': 596, + 'CSETv1:Embedded:yes': 592, + 'CSETv1_Annotator-1:Protected Characteristic:yes': 565, + 'CSETv1:Sector of Deployment:information and communication': 562, + 'CSETv1_Annotator-2:Special Interest Intangible Harm:no': 551, + 'CSETv1_Annotator-2:Physical Objects:no': 550, + 'CSETv1_Annotator-1:Autonomy Level:Autonomy3': 546, + 'CSETv0:Near Miss:Harm caused': 537, + 'CSETv1_Annotator-2:Clear link to Technology:yes': 522, + 'CSETv1_Annotator-1:Location Region:Global': 519, + 'CSETv0:Problem Nature:Robustness': 507, + 'CSETv1:Protected Characteristic:yes': 505, + 'CSETv0:Problem Nature:Specification': 473, + 'CSETv0:Near Miss:Unclear/unknown': 461, + 'CSETv1_Annotator-2:Location Region:North America': 454, + 'CSETv0:Level of Autonomy:High': 452, + 'CSETv1:Data Inputs:text': 443, + 'CSETv1_Annotator-2:Location Country (two letters):US': 441, + 'CSETv1_Annotator-3:Producer Test in Controlled Conditions:no': 435, + 'CSETv1_Annotator-3:Involving Minor:no': 427, + 'CSETv1_Annotator-3:User Test in Controlled Conditions:no': 424, + 'CSETv1_Annotator-2:Tangible Harm:no tangible harm, near-miss, or issue': 422, + 'CSETv1_Annotator-3:Producer Test in Operational Conditions:no': 420, + 'CSETv1_Annotator-3:Multiple AI Interaction:no': 418, + 'CSETv1_Annotator-2:Clear link to Technology:no': 414, + 'CSETv0:Level of Autonomy:Medium': 413, + 'CSETv1_Annotator-3:Report, Test, or Study of data:no': 412, + 'CSETv1_Annotator-3:Deployed:yes': 409, + 'CSETv1_Annotator-2:Physical Objects:yes': 404, + 'CSETv1:Location Region:Global': 402, + 'CSETv1_Annotator-3:Detrimental Content:no': 394, + 'CSETv1_Annotator-3:Entertainment Industry:no': 391, + 'CSETv1_Annotator-1:Date of Incident Year:2017': 389, + 'CSETv1_Annotator-2:Special Interest Intangible Harm:yes': 389, + 'CSETv1_Annotator-3:Impact on Critical Services:no': 388, + 'CSETv1_Annotator-3:Intentional Harm:No. Not intentionally designed to perform harm': 388, + 'CSETv1_Annotator-2:Tangible Harm:tangible harm definitively occurred': 383, + 'CSETv1:Date of Incident Year:2017': 382, + 'CSETv1_Annotator-1:AI Harm Level:AI tangible harm event': 381, + 'CSETv1_Annotator-2:Autonomy Level:Autonomy1': 379, + 'CSETv1:AI Harm Level:AI tangible harm event': 373, + 'CSETv1:Autonomy Level:Autonomy3': 370, + 'CSETv1_Annotator-1:Autonomy Level:Autonomy2': 369, + 'CSETv1_Annotator-1:Sector of Deployment:transportation and storage': 368, + 'CSETv1_Annotator-1:Harm Distribution Basis:race': 364, + 'CSETv0:Location:Global': 355, + 'CSETv0:Sector of Deployment:Information and communication': 352, + 'CSETv1_Annotator-3:AI System:yes': 350, + 'CSETv1:Harm Distribution Basis:race': 346, + 'CSETv1_Annotator-3:Public Sector Deployment:no': 346, + 'CSETv1_Annotator-3:User Test in Operational Conditions:no': 342, + 'CSETv1_Annotator-1:Clear link to Technology:no': 339, + 'CSETv1_Annotator-2:Annotator’s AI special interest intangible harm assessment:yes': 333, + 'CSETv1_Annotator-3:Rights Violation:no': 322, + 'CSETv1_Annotator-3:Clear link to Technology:yes': 305, + 'CSETv1_Annotator-1:Public Sector Deployment:yes': 304, + 'CSETv1_Annotator-2:Embedded:yes': 301, + 'CSETv0:Problem Nature:Unknown/unclear': 292, + 'CSETv1:Data Inputs:images': 286, + 'CSETv1_Annotator-1:AI System:no': 286, + 'CSETv1_Annotator-1:Data Inputs:images': 283, + 'CSETv1_Annotator-1:User Test in Operational Conditions:yes': 283, + 'CSETv0:Harm Type:Harm to physical health/safety': 282, + 'CSETv1_Annotator-2:Autonomy Level:Autonomy2': 281, + 'CSETv1_Annotator-1:Detrimental Content:yes': 279, + 'CSETv1:Sector of Deployment:transportation and storage': 273, + 'CSETv1:AI System:no': 269, + 'CSETv1_Annotator-2:Sector of Deployment:information and communication': 269, + 'CSETv1:Date of Incident Year:2016': 268, + 'CSETv1_Annotator-2:Date of Incident Year:2017': 267, + 'CSETv0:Severity:Minor': 262, + 'CSETv1:Detrimental Content:yes': 262, + 'CSETv1:Autonomy Level:Autonomy2': 261, + 'CSETv1_Annotator-2:Protected Characteristic:yes': 261, + 'CSETv1_Annotator-2:Data Inputs:Text': 260, + 'CSETv1_Annotator-3:Protected Characteristic:no': 258, + 'CSETv1_Annotator-3:Physical Objects:no': 256, + 'CSETv1_Annotator-1:Sector of Deployment:Arts, entertainment and recreation': 252, + 'CSETv1_Annotator-3:Harm Distribution Basis:none': 252, + 'CSETv1:Public Sector Deployment:yes': 249, + 'CSETv1_Annotator-1:Harm Distribution Basis:sex': 243, + 'CSETv0:Named Entities:Google': 242, + 'CSETv1_Annotator-3:Special Interest Intangible Harm:no': 241, + 'CSETv0:Harm Type:Harm to social or political systems': 239, + 'CSETv1_Annotator-1:Data Inputs:sensor data': 239, + 'CSETv0:Technology Purveyor:Google': 238, + 'CSETv0:Harm Distribution Basis:Race': 235, + 'CSETv1:AI tools and methods:natural language processing': 235, + 'CSETv1:Data Inputs:video': 234, + 'CSETv1_Annotator-1:Entertainment Industry:yes': 229, + 'CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:no': 226, + 'CSETv0:System Developer:Google': 224, + 'CSETv1_Annotator-1:Data Inputs:text': 223, + 'CSETv1_Annotator-1:Date of Incident Year:2016': 223, + 'CSETv1_Annotator-1:Date of Incident Year:2018': 222, + 'CSETv0:AI Techniques:Machine learning': 221, + 'CSETv1_Annotator-2:Autonomy Level:Autonomy3': 221, + 'CSETv0:Harm Type:Financial harm': 220, + 'CSETv0:Near Miss:Near miss': 219, + 'CSETv1_Annotator-3:Autonomy Level:Autonomy1': 219, + 'CSETv1_Annotator-1:Sector of Deployment:law enforcement': 218, + 'CSETv1:AI tools and methods:computer vision': 217, + 'CSETv1_Annotator-1:Data Inputs:video input': 216, + 'CSETv0:Physical System:Vehicle/mobile robot': 215, + 'CSETv1_Annotator-2:AI System:no': 215, + 'CSETv1_Annotator-3:Embedded:yes': 213, + 'CSETv1:User Test in Operational Conditions:yes': 211, + 'CSETv1_Annotator-3:Embedded:no': 211, + 'CSETv0:Problem Nature:Assurance': 207, + 'CSETv1_Annotator-2:Location Region:Global': 205, + 'CSETv1_Annotator-3:AI Harm Level:none': 204, + 'CSETv0:Harm Type:Psychological harm': 202, + 'CSETv1_Annotator-1:Tangible Harm:imminent risk of tangible harm (near miss) did occur': 201, + 'CSETv1:Harm Distribution Basis:sex': 198, + 'CSETv0:Intent:Unclear': 197, + 'CSETv1_Annotator-2:AI tools and methods:Natural Language Processesing': 196, + 'CSETv0:Sector of Deployment:Transportation and storage': 191, + 'CSETv1:Sector of Deployment:Arts, entertainment and recreation': 190, + 'CSETv1_Annotator-3:Location Region:North America': 187, + 'CSETv1_Annotator-3:Tangible Harm:tangible harm definitively occurred': 187, + 'CSETv0:AI Applications:Facial recognition': 185, + 'CSETv1_Annotator-3:Location Country (two letters):US': 185, + 'CSETv1_Annotator-1:Location Region:Europe': 183, + 'CSETv1_Annotator-2:Date of Incident Year:2016': 183, + 'CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:yes': 183, + 'CSETv1:Date of Incident Year:2018': 181, + 'CSETv1:Tangible Harm:imminent risk of tangible harm (near miss) did occur': 180, + 'CSETv1_Annotator-1:Producer Test in Operational Conditions:yes': 180, + 'CSETv1_Annotator-2:User Test in Operational Conditions:yes': 180, + 'CSETv1_Annotator-3:Physical Objects:yes': 180, + 'CSETv1_Annotator-3:Special Interest Intangible Harm:yes': 180, + 'CSETv0:Harm Type:Harm to civil liberties': 179, + 'CSETv1:Location Region:Asia': 177, + 'CSETv0:Nature of End User:Expert': 176, + 'CSETv1_Annotator-3:Sector of Deployment:information and communication': 176, + 'CSETv1_Annotator-2:Harm Distribution Basis:race': 174, + 'CSETv1_Annotator-3:Protected Characteristic:yes': 170, + 'CSETv1:Rights Violation:yes': 168, + 'CSETv1_Annotator-1:Location Region:Asia': 163, + 'CSETv1_Annotator-1:Involving Minor:yes': 157, + 'CSETv0:Sector of Deployment:Arts, entertainment and recreation': 156, + 'CSETv1_Annotator-3:AI Harm Level:AI tangible harm event': 156, + 'CSETv1_Annotator-1:AI Harm Level:AI tangible harm near-miss': 155, + 'CSETv1_Annotator-1:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred': 155, + 'CSETv0:Sector of Deployment:Public administration and defence': 154, + 'CSETv1:Sector of Deployment:wholesale and retail trade': 154, + 'CSETv1_Annotator-1:Data Inputs:camera input': 154, + 'CSETv1_Annotator-1:Sector of Deployment:wholesale and retail trade': 154, + 'CSETv1:Location Region:Europe': 152, + 'CSETv1_Annotator-2:Data Inputs:video': 150, + 'CSETv1_Annotator-3:Tangible Harm:no tangible harm, near-miss, or issue': 149, + 'CSETv0:AI Applications:autonomous driving': 147, + 'CSETv1:Data Inputs:sensor data': 147, + 'CSETv1_Annotator-2:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred': 147, + 'CSETv1_Annotator-1:Date of Incident Year:2023': 145, + 'CSETv0:Harm Distribution Basis:Sex': 144, + 'CSETv1:Data Inputs:audio': 143, + 'CSETv1:Data Inputs:lidar': 143, + 'CSETv1_Annotator-1:Entertainment Industry:maybe': 142, + 'CSETv0:Public Sector Deployment:true': 141, + 'CSETv1_Annotator-3:Date of Incident Year:2017': 141, + 'CSETv1:Entertainment Industry:yes': 140, + 'CSETv0:AI Techniques:Facial recognition': 139, + 'CSETv1:Infrastructure Sectors:transportation': 139, + 'CSETv1_Annotator-3:Location Region:Global': 138, + 'CSETv0:Named Entities:Amazon': 136, + 'CSETv1_Annotator-1:AI tools and methods:natural language processing': 135, + 'GMF:Potential AI Technical Failure:Generalization Failure': 135, + 'CSETv1_Annotator-1:Data Inputs:facial images': 134, + 'CSETv0:AI Techniques:Natural language processing': 133, + 'CSETv0:Harm Type:Harm to physical property': 132, + 'CSETv1_Annotator-1:AI Harm Level:AI tangible harm issue': 132, + 'CSETv1:AI Task:facial recognition': 131, + 'CSETv1:Data Inputs:radar': 131, + 'CSETv0:Technology Purveyor:Amazon': 129, + 'CSETv1:Producer Test in Operational Conditions:yes': 129, + 'CSETv1_Annotator-1:Date of Incident Year:2020': 129, + 'CSETv1_Annotator-2:AI tools and methods:computer vision': 129, + 'CSETv1_Annotator-1:Date of Incident Year:2021': 128, + 'CSETv1_Annotator-1:Date of Incident Year:2022': 128, + 'CSETv1_Annotator-2:Public Sector Deployment:yes': 128, + 'CSETv1_Annotator-2:AI Harm Level:AI tangible harm event': 126, + 'CSETv1_Annotator-2:Location Region:Asia': 126, + 'CSETv1:Sector of Deployment:administrative and support service activities': 125, + 'CSETv1_Annotator-3:Autonomy Level:Autonomy3': 125, + 'CSETv1_Annotator-1:Rights Violation:yes': 124, + 'CSETv1_Annotator-1:Sector of Deployment:public administration': 123, + 'CSETv0:Infrastructure Sectors:Transportation': 122, + 'CSETv0:Level of Autonomy:Low': 122, + 'CSETv1:Sector of Deployment:public administration': 117, + 'CSETv1_Annotator-1:AI Task:facial recognition': 117, + 'CSETv1_Annotator-2:Harm Distribution Basis:sex': 116, + 'CSETv1_Annotator-1:Harm Distribution Basis:nation of origin, citizenship, immigrant status': 115, + 'CSETv1_Annotator-1:Sector of Deployment:professional, scientific and technical activities': 115, + 'CSETv0:Severity:Unclear/unknown': 114, + 'CSETv1:Sector of Deployment:law enforcement': 114, + 'CSETv1_Annotator-1:AI Task:navigation': 114, + 'CSETv0:AI Applications:decision support': 113, + 'CSETv0:System Developer:Amazon': 112, + 'CSETv1:AI Task:navigation': 112, + 'CSETv1_Annotator-2:Involving Minor:yes': 111, + 'CSETv1_Annotator-2:Detrimental Content:yes': 110, + 'CSETv0:Physical System:Consumer device': 108, + 'CSETv1_Annotator-1:Clear link to Technology:maybe': 108, + 'CSETv1_Annotator-1:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm': 107, + 'CSETv1_Annotator-2:Sector of Deployment:transportation and storage': 107, + 'CSETv1_Annotator-3:AI tools and methods:unclear': 107, + 'CSETv0:AI Applications:recommendation engine': 106, + 'CSETv1:Impact on Critical Services:yes': 106, + 'CSETv1:Involving Minor:yes': 106, + 'CSETv1_Annotator-3:Data Inputs:text': 106, + 'CSETv0:Severity:Moderate': 105, + 'CSETv1:Clear link to Technology:no': 105, + 'CSETv1:AI Harm Level:AI tangible harm near-miss': 104, + 'CSETv1:Autonomy Level:unclear': 104, + 'CSETv1_Annotator-1:AI tools and methods:facial recognition': 104, + 'CSETv1_Annotator-2:Data Inputs:images': 104, + 'CSETv0:Lives Lost:true': 103, + 'CSETv0:Technology Purveyor:Tesla': 103, + 'CSETv1:Deployed:no': 103, + 'CSETv1_Annotator-1:AI Task:autonomous navigation': 103, + 'CSETv0:Severity:Severe': 102, + 'CSETv1_Annotator-1:Autonomy Level:unclear': 102, + 'CSETv1_Annotator-1:Date of Incident Year:2014': 102, + 'CSETv1_Annotator-2:Sector of Deployment:wholesale and retail trade': 102, + 'CSETv1_Annotator-2:Location Region:Europe': 101, + 'CSETv1:Data Inputs:facial images': 100, + 'CSETv1:Harm Distribution Basis:nation of origin, citizenship, immigrant status': 100, + 'CSETv1_Annotator-3:Harm Distribution Basis:race': 100, + 'CSETv1_Annotator-1:AI Task:text generation': 99, + 'CSETv1_Annotator-2:Sector of Deployment:administrative and support service activities': 99, + 'CSETv1:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred': 97, + 'CSETv1_Annotator-1:AI Task:semi-autonomous navigation': 96, + 'CSETv1:Data Inputs:camera input': 95, + 'CSETv1:Date of Incident Year:2014': 95, + 'CSETv1_Annotator-2:Date of Incident Year:2015': 95, + 'CSETv1_Annotator-2:Rights Violation:yes': 94, + 'GMF:Potential AI Technical Failure:Context Misidentification': 94, + 'CSETv1:AI System:maybe': 92, + 'CSETv1:Sector of Deployment:human health and social work activities': 92, + 'CSETv1_Annotator-1:Detrimental Content:maybe': 91, + 'CSETv1_Annotator-3:User Test in Operational Conditions:yes': 91, + 'CSETv0:AI Applications:image classification': 90, + 'CSETv0:Sector of Deployment:Administrative and support service activities': 90, + 'CSETv1_Annotator-1:Harm Distribution Basis:religion': 90, + 'CSETv1_Annotator-1:Location Region:Oceania': 90, + 'CSETv1_Annotator-3:Rights Violation:yes': 90, + 'CSETv1_Annotator-2:Data Inputs:audio': 89, + 'CSETv0:Level of Autonomy:Unclear/unknown': 88, + 'CSETv1:Harm Distribution Basis:religion': 88, + 'CSETv1_Annotator-1:AI Task:content generation': 88, + 'CSETv1_Annotator-1:AI System:maybe': 86, + 'CSETv1_Annotator-1:Data Inputs:radar': 86, + 'CSETv0:AI Applications:Natural language processing': 85, + 'CSETv0:Intent:Deliberate or expected': 85, + 'CSETv1_Annotator-1:Harm Distribution Basis:financial means': 84, + 'CSETv1:Date of Incident Year:2020': 83, + 'CSETv1_Annotator-2:Report, Test, or Study of data:yes': 83, + 'CSETv1:Date of Incident Year:2021': 82, + 'CSETv1:Rights Violation:maybe': 82, + 'CSETv1_Annotator-2:AI Harm Level:AI tangible harm issue': 82, + 'CSETv1_Annotator-3:Clear link to Technology:no': 82, + 'CSETv1_Annotator-3:Sector of Deployment:law enforcement': 82, + 'CSETv1_Annotator-1:Producer Test in Controlled Conditions:yes': 80, + 'CSETv0:Named Entities:Tesla': 79, + 'CSETv1_Annotator-1:AI Task:chat bot': 79, + 'CSETv1_Annotator-1:Data Inputs:spatial data': 79, + 'CSETv1_Annotator-1:Deployed:no': 79, + 'CSETv0:AI Applications:image recognition': 78, + 'CSETv1:Date of Incident Year:2015': 78, + 'CSETv1_Annotator-1:Data Inputs:search queries': 78, + 'CSETv1_Annotator-3:Public Sector Deployment:yes': 78, + 'CSETv1:AI tools and methods:Natural Language Processesing': 77, + 'CSETv1:Date of Incident Year:2023': 77, + 'CSETv1_Annotator-1:AI tools and methods:machine learning': 77, + 'CSETv1_Annotator-1:Harm Distribution Basis:sexual orientation or gender identity': 77, + 'CSETv0:AI Techniques:Unclear': 76, + 'CSETv1:Lives Lost:1': 76, + 'CSETv1_Annotator-2:Deployed:no': 76, + 'CSETv0:AI Techniques:Tesla Autopilot': 75, + 'CSETv0:System Developer:Tesla': 75, + 'CSETv1:Infrastructure Sectors:healthcare and public health': 75, + 'CSETv1_Annotator-1:Deployed:maybe': 75, + 'CSETv1_Annotator-2:Date of Incident Year:2018': 74, + 'CSETv1_Annotator-2:Infrastructure Sectors:transportation': 74, + 'CSETv1:Harm Distribution Basis:sexual orientation or gender identity': 73, + 'CSETv1_Annotator-1:Data Inputs:radar input': 73, + 'CSETv1_Annotator-2:Autonomy Level:unclear': 73, + 'CSETv1:Sector of Deployment:professional, scientific and technical activities': 71, + 'CSETv1_Annotator-1:Date of Incident Year:2015': 71, + 'CSETv1_Annotator-1:Sector of Deployment:financial and insurance activities': 71, + 'CSETv1:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm': 70, + 'CSETv1_Annotator-3:Sector of Deployment:transportation and storage': 70, + 'CSETv0:Relevant AI functions:Unclear': 69, + 'CSETv1:Producer Test in Controlled Conditions:yes': 69, + 'CSETv1:Sector of Deployment:financial and insurance activities': 69, + 'GMF:Known AI Technology:Face Detection': 69, + 'CSETv1:Data Inputs:search queries': 68, + 'CSETv1_Annotator-1:Data Inputs:voice input': 68, + 'CSETv1_Annotator-1:Sector of Deployment:human health and social work activities': 68, + 'CSETv1_Annotator-2:Infrastructure Sectors:healthcare and public health': 68, + 'CSETv1_Annotator-2:Sector of Deployment:human health and social work activities': 68, + 'CSETv1:Physical System Type:Manufacturing Robot': 67, + 'CSETv1:Sector of Deployment:manufacturing': 67, + 'CSETv1_Annotator-2:Sector of Deployment:manufacturing': 67, + 'CSETv1:AI Harm Level:AI tangible harm issue': 66, + 'CSETv1:Injuries:1': 66, + 'CSETv1_Annotator-1:Data Inputs:user queries': 66, + 'CSETv1_Annotator-1:Lives Lost:1': 66, + 'CSETv1_Annotator-1:Location Country (two letters):AU': 66, + 'CSETv1_Annotator-2:Producer Test in Controlled Conditions:yes': 66, + 'CSETv1_Annotator-3:Data Inputs:video': 65, + 'CSETv0:Named Entities:Microsoft': 64, + 'CSETv1:Data Inputs:speech': 64, + 'CSETv1:Location Region:Oceania': 64, + 'CSETv1_Annotator-1:AI Task:self-driving': 64, + 'CSETv1_Annotator-1:Data Inputs:geospatial data': 64, + 'CSETv1_Annotator-2:AI System:maybe': 64, + 'CSETv1_Annotator-2:Date of Incident Year:2023': 64, + 'CSETv1_Annotator-2:Sector of Deployment:financial and insurance activities': 64, + 'CSETv1:Operating Conditions:operationally representative': 63, + 'CSETv1:Report, Test, or Study of data:yes': 63, + 'CSETv0:AI Applications:self-driving vehicle': 62, + 'CSETv0:AI Techniques:Deep learning': 62, + 'CSETv0:Named Entities:Knightscope': 62, + 'CSETv0:Named Entities:Knightscope K5': 62, + 'CSETv0:System Developer:Knightscope': 62, + 'CSETv0:Technology Purveyor:Knightscope': 62, + 'CSETv1:AI Task:patrolling': 62, + 'CSETv1_Annotator-1:AI Task:virtual assistant technology': 62, + 'CSETv1_Annotator-1:Physical System Type:rocket/egg shaped, 300 pound, 5 ft. tall security robot': 62, + 'CSETv1_Annotator-2:Data Inputs:lidar': 62, + 'CSETv1_Annotator-2:Location Region:Oceania': 62, + 'CSETv1_Annotator-3:Location Region:Asia': 62, + 'CSETv1:Data Inputs:video input': 61, + 'CSETv1:Data Inputs:voice': 61, + 'CSETv1_Annotator-1:Injuries:1': 61, + 'CSETv1_Annotator-2:Data Inputs:voice': 61, + 'CSETv0:Sector of Deployment:Financial and insurance activities': 60, + 'CSETv1_Annotator-1:Data Inputs:lidar': 60, + 'CSETv1_Annotator-1:Data Inputs:user prompts': 60, + 'CSETv1_Annotator-2:Intentional Harm:Yes. Intentionally designed to perform harm and did create intended harm': 60, + 'CSETv1_Annotator-2:Sector of Deployment:law enforcement': 60, + 'CSETv0:Harm Distribution Basis:National origin or immigrant status': 59, + 'CSETv1:Location Country (two letters):RU': 59, + 'CSETv1:Physical System Type:none': 59, + 'CSETv1_Annotator-1:AI tools and methods:voice recognition': 59, + 'CSETv1_Annotator-1:Data Inputs:speech': 59, + 'CSETv1_Annotator-1:Report, Test, or Study of data:maybe': 59, + 'CSETv1_Annotator-2:Physical System Type:none': 59, + 'GMF:Known AI Technical Failure:Unsafe Exposure or Access': 59, + 'CSETv1:AI Task:virtual assistant technology': 58, + 'CSETv1:Data Inputs:photos': 58, + 'CSETv0:AI Techniques:environmental sensing': 57, + 'CSETv0:Named Entities:Stanford Shopping Center': 57, + 'CSETv1:AI Task:security': 57, + 'CSETv1:Data Inputs:road data': 57, + 'CSETv1:Physical System Type:rocket/egg shaped, 300 pound, 5 ft. tall security robot': 57, + 'CSETv1_Annotator-1:AI Task:patrolling': 57, + 'CSETv1_Annotator-1:AI Task:security': 57, + 'CSETv1_Annotator-2:AI tools and methods:Text to Speach': 57, + 'CSETv1_Annotator-2:Data Inputs:road data': 57, + 'CSETv1_Annotator-2:Operating Conditions:operationally representative': 57, + 'CSETv1_Annotator-1:Location Country (two letters):CN': 56, + 'CSETv1_Annotator-3:Date of Incident Year:2016': 56, + 'GMF:Potential AI Technical Failure:Dataset Imbalance': 56, + 'CSETv0:AI Techniques:open-source': 55, + 'CSETv0:Technology Purveyor:Microsoft': 55, + 'CSETv1:AI Task:object detection': 55, + 'CSETv1:AI tools and methods:facial recognition': 55, + 'CSETv1:AI tools and methods:voice recognition': 55, + 'CSETv1_Annotator-1:Physical Objects:maybe': 55, + 'CSETv1_Annotator-1:Report, Test, or Study of data:yes': 55, + 'CSETv1_Annotator-1:Rights Violation:maybe': 55, + 'CSETv1_Annotator-1:Sector of Deployment:manufacturing': 55, + 'CSETv1_Annotator-2:AI tools and methods:Audio Transcription': 55, + 'CSETv1_Annotator-2:Physical System Type:Manufacturing Robot': 55, + 'GMF:Known AI Technical Failure:Distributional Bias': 55, + 'GMF:Known AI Technology:Recurrent Neural Network': 55, + 'CSETv1:AI Task:autonomous driving': 54, + 'CSETv1:AI tools and methods:neural networks': 54, + 'CSETv1_Annotator-1:Data Inputs:audio inputs': 54, + 'GMF:Potential AI Technology:Regression': 54, + 'CSETv0:Harm Distribution Basis:Financial means': 53, + 'CSETv1:AI Task:self-driving': 53, + 'CSETv1:Date of Incident Year:2013': 53, + 'CSETv1:Location Country (two letters):': 53, + 'CSETv1_Annotator-1:AI Task:image identification': 53, + 'CSETv1_Annotator-2:Infrastructure Sectors:communications': 53, + 'GMF:Known AI Technology:Distributional Learning': 53, + 'CSETv0:AI Techniques:image classification': 52, + 'CSETv1:Clear link to Technology:maybe': 52, + 'CSETv1:Data Inputs:spatial data': 52, + 'CSETv1_Annotator-1:Date of Incident Year:2019': 52, + 'CSETv1_Annotator-1:Sector of Deployment:Education': 52, + 'GMF:Potential AI Technical Failure:Underspecification': 52, + 'CSETv0:Location:Palo Alto, CA': 51, + 'CSETv0:Named Entities:Apple': 51, + 'CSETv0:Named Entities:Joshua Brown': 51, + 'CSETv0:Technology Purveyor:Apple': 51, + 'CSETv1:AI Task:object recognition': 51, + 'CSETv1:AI tools and methods:audio transcription': 51, + 'CSETv1:AI tools and methods:machine learning': 51, + 'CSETv1:Data Inputs:infrared camera': 51, + 'CSETv1:Data Inputs:personal data': 51, + 'CSETv1_Annotator-1:Data Inputs:infrared camera': 51, + 'CSETv1_Annotator-2:Impact on Critical Services:yes': 51, + 'CSETv1_Annotator-2:Physical System Type:Amazon Alexa': 51, + 'CSETv1_Annotator-1:Special Interest Intangible Harm:maybe': 50, + 'GMF:Known AI Technology:Convolutional Neural Network': 50, + 'CSETv0:AI Applications:chatbot': 49, + 'CSETv0:Harm Distribution Basis:Religion': 49, + 'CSETv1:AI Task:chat bot': 49, + 'CSETv1_Annotator-1:AI Task:smart suggestions': 49, + 'CSETv1:Multiple AI Interaction:yes': 48, + 'CSETv1_Annotator-1:AI Harm Level:unclear': 48, + 'CSETv1_Annotator-1:Date of Incident Year:2013': 48, + 'CSETv1_Annotator-1:Infrastructure Sectors:transportation': 48, + 'CSETv1_Annotator-1:Multiple AI Interaction:yes': 48, + 'CSETv1_Annotator-3:AI Harm Level:unclear': 48, + 'CSETv1_Annotator-3:Data Inputs:audio': 48, + 'CSETv1_Annotator-3:Location Region:Europe': 48, + 'CSETv1:Data Inputs:image alt tags': 47, + 'CSETv1_Annotator-1:Data Inputs:image alt tags': 47, + 'CSETv1_Annotator-2:AI Task:autonomous driving': 47, + 'CSETv1_Annotator-3:Impact on Critical Services:yes': 47, + 'CSETv0:AI Techniques:autonomous vehicle': 46, + 'CSETv1:AI Task:text generation': 46, + 'CSETv1_Annotator-1:Impact on Critical Services:yes': 46, + 'CSETv1_Annotator-2:Location Country (two letters):': 46, + 'CSETv0:System Developer:Apple': 45, + 'CSETv1:AI tools and methods:image mapping': 45, + 'CSETv1:AI tools and methods:point mapping': 45, + 'CSETv1:Data Inputs:dot projector': 45, + 'CSETv1:Location Country (two letters):CN': 45, + 'CSETv1:Physical System Type:Apple iPhone X': 45, + 'CSETv1_Annotator-1:AI Task:image generation': 45, + 'CSETv1_Annotator-1:AI tools and methods:facial reconstruction': 45, + 'CSETv1_Annotator-1:AI tools and methods:image mapping': 45, + 'CSETv1_Annotator-1:AI tools and methods:image reconstruction': 45, + 'CSETv1_Annotator-1:AI tools and methods:point mapping': 45, + 'CSETv1_Annotator-1:Data Inputs:dot projector': 45, + 'CSETv1_Annotator-1:Physical System Type:Apple iPhone X': 45, + 'CSETv1_Annotator-2:Date of Incident Year:2020': 45, + 'CSETv1_Annotator-2:Producer Test in Operational Conditions:yes': 45, + 'CSETv1_Annotator-3:AI System:no': 45, + 'CSETv1_Annotator-3:Data Inputs:radar': 45, + 'CSETv1_Annotator-3:Harm Distribution Basis:sex': 45, + 'CSETv1_Annotator-3:Harm Distribution Basis:sexual orientation or gender identity': 45, + 'GMF:Known AI Goal:Autonomous Driving': 45, + 'GMF:Potential AI Technical Failure:Unauthorized Data': 45, + 'GMF:Potential AI Technology:Image Segmentation': 45, + 'CSETv0:AI Applications:voice recognition': 44, + 'CSETv0:System Developer:Microsoft': 44, + 'CSETv1:AI Task:search optimization': 44, + 'CSETv1:Data Inputs:GPS': 44, + 'CSETv1:Harm Distribution Basis:financial means': 44, + 'CSETv1:Intentional Harm:unclear': 44, + 'CSETv1_Annotator-2:Entertainment Industry:yes': 44, + 'CSETv1_Annotator-3:Harm Distribution Basis:religion': 44, + 'GMF:Potential AI Technology:Classification': 44, + 'GMF:Potential AI Technology:Convolutional Neural Network': 44, + 'CSETv0:AI Techniques:language recognition': 43, + 'CSETv1:AI Task:autonomous navigation': 43, + 'CSETv1:Date of Incident Year:2011': 43, + 'CSETv1:Date of Incident Year:2019': 43, + 'CSETv1_Annotator-1:AI Task:content moderation': 43, + 'CSETv1_Annotator-1:AI Task:search optimization': 43, + 'CSETv1_Annotator-1:AI Task:search suggestion': 43, + 'CSETv1_Annotator-1:Location Country (two letters):DE': 43, + 'CSETv1_Annotator-2:AI Task:self driving': 43, + 'CSETv1_Annotator-2:Multiple AI Interaction:yes': 43, + 'GMF:Known AI Technology:Language Modeling': 43, + 'CSETv0:Location:Los Angeles, CA': 42, + 'CSETv1:AI tools and methods:unclear': 42, + 'CSETv1_Annotator-1:Date of Incident Year:2012': 42, + 'CSETv1_Annotator-1:Sector of Deployment:administrative and support service activities': 42, + 'CSETv1_Annotator-3:Detrimental Content:yes': 42, + 'CSETv0:Named Entities:China': 41, + 'CSETv1:AI Task:search suggestion': 41, + 'CSETv1:Date of Incident Year:2012': 41, + 'CSETv1:Entertainment Industry:maybe': 41, + 'CSETv1:Harm Distribution Basis:disability': 41, + 'CSETv1:Location Country (two letters):IN': 41, + 'CSETv1:Sector of Deployment:Education': 41, + 'CSETv1_Annotator-1:Data Inputs:age': 41, + 'CSETv1_Annotator-1:Harm Distribution Basis:disability': 41, + 'CSETv1_Annotator-1:Lives Lost:2': 41, + 'CSETv1_Annotator-2:Date of Incident Year:2014': 41, + 'CSETv1_Annotator-2:Location Country (two letters):IN': 41, + 'CSETv1_Annotator-3:AI System:maybe': 41, + 'CSETv1_Annotator-3:Autonomy Level:Autonomy2': 41, + 'GMF:Potential AI Technical Failure:Underfitting': 41, + 'CSETv0:AI Applications:environmental sensing': 40, + 'CSETv0:AI Techniques:virtual assistant': 40, + 'CSETv0:Harm Distribution Basis:Sexual orientation or gender identity': 40, + 'CSETv1:AI Task:Image search': 40, + 'CSETv1:Location Country (two letters):AU': 40, + 'CSETv1_Annotator-1:Harm Distribution Basis:ideology': 40, + 'CSETv1_Annotator-2:AI Task:Personal Assistant': 40, + 'CSETv1_Annotator-2:Date of Incident Year:2021': 40, + 'CSETv1_Annotator-3:Tangible Harm:imminent risk of tangible harm (near miss) did occur': 40, + 'CSETv0:AI Techniques:Amazon Alexa': 39, + 'CSETv0:Location:Australia': 39, + 'CSETv0:Named Entities:Alan Tudge': 39, + 'CSETv0:Named Entities:Centrelink': 39, + 'CSETv0:Named Entities:Centrelink Master Program': 39, + 'CSETv0:Named Entities:Department of Human Services': 39, + 'CSETv0:Named Entities:Richard Glenn': 39, + 'CSETv0:System Developer:Centrelink Master Program': 39, + 'CSETv0:Technology Purveyor:Centrelink Master Program': 39, + 'CSETv0:Technology Purveyor:Department of Human Services': 39, + 'CSETv1:Data Inputs:income data': 39, + 'CSETv1:Data Inputs:tabular': 39, + 'CSETv1:Data Inputs:tax reports': 39, + 'CSETv1:Detrimental Content:maybe': 39, + 'CSETv1_Annotator-1:Data Inputs:income data': 39, + 'CSETv1_Annotator-1:Data Inputs:laser input': 39, + 'CSETv1_Annotator-1:Data Inputs:tax reports': 39, + 'CSETv1_Annotator-1:Data Inputs:words': 39, + 'CSETv1_Annotator-1:Infrastructure Sectors:government facilities': 39, + 'CSETv1_Annotator-2:Data Inputs:amazon account': 39, + 'CSETv1_Annotator-2:Infrastructure Sectors:government facilities': 39, + 'CSETv1_Annotator-2:Lives Lost:1': 39, + 'CSETv1_Annotator-2:Location Country (two letters):AU': 39, + 'CSETv1_Annotator-3:Autonomy Level:unclear': 39, + 'CSETv1_Annotator-3:Data Inputs:lidar': 39, + 'CSETv1_Annotator-3:Sector of Deployment:wholesale and retail trade': 39, + 'GMF:Potential AI Technical Failure:Hardware Failure': 39, + 'GMF:Potential AI Technical Failure:Lack of Capability Control': 39, + 'GMF:Potential AI Technology:Ensemble Aggregation': 39, + 'CSETv0:AI Techniques:natural language processing model': 38, + 'CSETv1:AI Task:predict recidivism': 38, + 'CSETv1:AI tools and methods:object detection': 38, + 'CSETv1_Annotator-1:AI Task:predict recidivism': 38, + 'CSETv1_Annotator-1:AI tools and methods:computer vision': 38, + 'CSETv1_Annotator-2:AI tools and methods:Machine Learning': 38, + 'CSETv1_Annotator-2:Sector of Deployment:other': 38, + 'GMF:Potential AI Technical Failure:Data or Labelling Noise': 38, + 'CSETv0:AI Techniques:law enforcement algorithm': 37, + 'CSETv0:Infrastructure Sectors:Government facilities': 37, + 'CSETv1:AI Task:Rank Applicants': 37, + 'CSETv1:AI Task:translation': 37, + 'CSETv1:Data Inputs:Amazon account information': 37, + 'CSETv1:Deployed:maybe': 37, + 'CSETv1_Annotator-1:AI Task:search result ranking': 37, + 'CSETv1_Annotator-1:Data Inputs:137 questionnaire responses': 37, + 'CSETv1_Annotator-1:Data Inputs:Amazon account information': 37, + 'CSETv1_Annotator-2:AI Task:Rank Applicants': 37, + 'CSETv1_Annotator-3:Date of Incident Year:2020': 37, + 'GMF:Known AI Technical Failure:Misuse': 37, + 'GMF:Known AI Technology:Automatic Speech Recognition': 37, + 'CSETv0:AI Applications:language translation': 36, + 'CSETv1:AI Harm Level:unclear': 36, + 'CSETv1:Data Inputs:Facebook posts': 36, + 'CSETv1:Data Inputs:air quality': 36, + 'CSETv1:Date of Incident Year:2022': 36, + 'CSETv1:Infrastructure Sectors:financial services': 36, + 'CSETv1_Annotator-1:AI Task:transportation': 36, + 'CSETv1_Annotator-1:Location Country (two letters):IN': 36, + 'CSETv1_Annotator-1:Location Country (two letters):RU': 36, + 'CSETv1_Annotator-1:Physical System Type:Tesla Model S': 36, + 'CSETv1_Annotator-3:Entertainment Industry:yes': 36, + 'GMF:Known AI Goal:Face Recognition': 36, + 'CSETv0:AI Applications:Interpreting traffic patterns': 35, + 'CSETv0:AI Applications:image processing': 35, + 'CSETv0:Harm Distribution Basis:Geography': 35, + 'CSETv0:Infrastructure Sectors:Information technology': 35, + 'CSETv0:Named Entities:San Diego TV': 35, + 'CSETv1:AI Task:prediction': 35, + 'CSETv1:AI Task:self driving': 35, + 'CSETv1:AI Task:semi-autonomous navigation': 35, + 'CSETv1:AI Task:surveillance': 35, + 'CSETv1:Data Inputs:Crime Records': 35, + 'CSETv1:Harm Distribution Basis:geography': 35, + 'CSETv1:Location Country (two letters):IE': 35, + 'CSETv1:Physical System Type:Amazon Alexa Echo Dot speaker': 35, + 'CSETv1:Physical System Type:Volvo XC90 SUV': 35, + 'CSETv1_Annotator-1:AI Task:image modification': 35, + 'CSETv1_Annotator-1:AI tools and methods:audio transcription': 35, + 'CSETv1_Annotator-1:Location Country (two letters):IE': 35, + 'CSETv1_Annotator-1:Physical System Type:Amazon Alexa Echo Dot speaker': 35, + 'CSETv1_Annotator-1:Sector of Deployment:defense': 35, + 'CSETv1_Annotator-2:AI Harm Level:AI tangible harm near-miss': 35, + 'CSETv1_Annotator-2:Deployed:maybe': 35, + 'CSETv1_Annotator-3:Data Inputs:face image': 35, + 'CSETv1_Annotator-3:Date of Incident Year:2021': 35, + 'GMF:Known AI Goal:AI Voice Assistant': 35, + 'GMF:Known AI Technology:Acoustic Fingerprint': 35, + 'GMF:Known AI Technology:Neural Network': 35, + 'GMF:Potential AI Technical Failure:Inadequate Anonymization': 35, + 'GMF:Potential AI Technology:Distributional Learning': 35, + 'CSETv0:AI Techniques:radar': 34, + 'CSETv0:Harm Type:Harm to intangible property': 34, + 'CSETv0:Sector of Deployment:Professional, scientific and technical activities': 34, + 'CSETv1:AI Task:Driving': 34, + 'CSETv1_Annotator-1:Data Inputs:photos': 34, + 'CSETv1_Annotator-1:Protected Characteristic:maybe': 34, + 'CSETv1_Annotator-1:Public Sector Deployment:maybe': 34, + 'CSETv1_Annotator-2:AI Task:recommender': 34, + 'CSETv1_Annotator-2:Sector of Deployment:Arts, entertainment and recreation': 34, + 'GMF:Potential AI Technical Failure:Overfitting': 34, + 'CSETv0:Location:Edinburgh, Scotland': 33, + 'CSETv0:Named Entities:Edinburgh': 33, + 'CSETv1:AI Task:resume screening': 33, + 'CSETv1:Data Inputs:Resume': 33, + 'CSETv1:Data Inputs:traffic': 33, + 'CSETv1:Harm Distribution Basis:ideology': 33, + 'CSETv1_Annotator-1:AI Task:obstacle avoidance': 33, + 'CSETv1_Annotator-1:AI Task:recruitment': 33, + 'CSETv1_Annotator-1:AI Task:resume screening': 33, + 'CSETv1_Annotator-1:Data Inputs:resumes': 33, + 'CSETv1_Annotator-2:AI tools and methods:Natural language processing': 33, + 'CSETv1_Annotator-2:Data Inputs:Resume': 33, + 'CSETv1_Annotator-2:Data Inputs:personal data': 33, + 'CSETv1_Annotator-2:Data Inputs:photos': 33, + 'CSETv1_Annotator-2:Data Inputs:traffic': 33, + 'CSETv1_Annotator-2:Date of Incident Year:2013': 33, + 'CSETv1_Annotator-2:Infrastructure Sectors:unclear': 33, + 'CSETv1_Annotator-2:Location Country (two letters):IE': 33, + 'CSETv1_Annotator-2:Rights Violation:maybe': 33, + 'CSETv1_Annotator-3:Tangible Harm:unclear': 33, + 'GMF:Known AI Technology:Generative Adversarial Network': 33, + 'GMF:Potential AI Technical Failure:Misinformation Generation Hazard': 33, + 'CSETv0:AI Applications:content filtering': 32, + 'CSETv0:AI Applications:environment sensing': 32, + 'CSETv0:AI Applications:self-driving': 32, + 'CSETv0:Named Entities:Facebook': 32, + 'CSETv0:System Developer:Facebook': 32, + 'CSETv0:Technology Purveyor:Facebook': 32, + 'CSETv1:AI tools and methods:human language technology': 32, + 'CSETv1:Location Country (two letters):DE': 32, + 'CSETv1_Annotator-1:AI Task:deepfake video generation': 32, + 'CSETv1:AI Task:content generation': 31, + 'CSETv1:AI Task:deepfake video generation': 31, + 'CSETv1:AI tools and methods:search engine optimization': 31, + 'CSETv1:Data Inputs:age': 31, + 'CSETv1:Data Inputs:personal information': 31, + 'CSETv1:Infrastructure Sectors:defense-industrial base': 31, + 'CSETv1:Sector of Deployment:defense': 31, + 'CSETv1_Annotator-1:AI Task:image search': 31, + 'CSETv1_Annotator-1:AI tools and methods:neural networks': 31, + 'CSETv1_Annotator-1:AI tools and methods:search engine optimization': 31, + 'CSETv1_Annotator-1:Sector of Deployment:other': 31, + 'CSETv1_Annotator-2:Location Country (two letters):DE': 31, + 'GMF:Known AI Goal:Autonomous Drones': 31, + 'CSETv0:AI Applications:forecasting': 30, + 'CSETv0:AI Applications:stock trading': 30, + 'CSETv0:AI Techniques:stock market algorithm': 30, + 'CSETv0:Financial Cost:Short term: $1 trillion unclear of long-term impact': 30, + 'CSETv0:Infrastructure Sectors:Financial services': 30, + 'CSETv0:Location:UK/USA': 30, + 'CSETv0:Location:Washington, D.C.': 30, + 'CSETv0:Named Entities:Chicago Merchant Exchange': 30, + 'CSETv0:Named Entities:Dow Jones Industrial Index': 30, + 'CSETv0:Named Entities:Electronic Privacy Information Center (EPIC)': 30, + 'CSETv0:Named Entities:Mountain View': 30, + 'CSETv0:Named Entities:Navinder Singh Saroa': 30, + 'CSETv0:Technology Purveyor:Navinder Singh Sarao': 30, + 'CSETv1:AI Task:autocomplete': 30, + 'CSETv1:AI Task:content moderation': 30, + 'CSETv1:AI Task:recommender': 30, + 'CSETv1:Data Inputs:radar input': 30, + 'CSETv1:Data Inputs:stock data': 30, + 'CSETv1:Data Inputs:stock orders': 30, + 'CSETv1:Data Inputs:stock prices': 30, + 'CSETv1:Data Inputs:thermal imaging': 30, + 'CSETv1:Data Inputs:words': 30, + 'CSETv1:Date of Incident Year:2010': 30, + 'CSETv1:Harm Distribution Basis:age': 30, + 'CSETv1_Annotator-1:AI Task:surveillance': 30, + 'CSETv1_Annotator-1:Data Inputs:Youtube videos': 30, + 'CSETv1_Annotator-1:Data Inputs:air quality data': 30, + 'CSETv1_Annotator-1:Data Inputs:personal information': 30, + 'CSETv1_Annotator-1:Data Inputs:stock orders': 30, + 'CSETv1_Annotator-1:Data Inputs:stock prices': 30, + 'CSETv1_Annotator-1:Data Inputs:thermal imaging': 30, + 'CSETv1_Annotator-1:Date of Incident Year:2010': 30, + 'CSETv1_Annotator-1:Infrastructure Sectors:financial services': 30, + 'CSETv1_Annotator-2:Data Inputs:stock data': 30, + 'CSETv1_Annotator-2:Date of Incident Year:2010': 30, + 'CSETv1_Annotator-3:AI Task:Patrolling': 30, + 'CSETv1_Annotator-3:Data Inputs:Facebook posts': 30, + 'CSETv1_Annotator-3:Data Inputs:Unclear': 30, + 'CSETv1_Annotator-3:Data Inputs:air quality': 30, + 'CSETv1_Annotator-3:Data Inputs:thermal imaging': 30, + 'CSETv1_Annotator-3:Physical System Type:autonomous security robot': 30, + 'GMF:Known AI Goal:Automatic Stock Trading': 30, + 'GMF:Potential AI Technical Failure:Gaming Vulnerability': 30, + 'CSETv0:AI Applications:Content curation': 29, + 'CSETv0:AI Techniques:GANs': 29, + 'CSETv0:AI Techniques:RNNs': 29, + 'CSETv0:Harm Type:Other:Reputational harm/social harm (libel and defamation)': 29, + 'CSETv0:Location:Delhi, India': 29, + 'CSETv0:Location:Unknown': 29, + 'CSETv0:Location:Williston, FL': 29, + 'CSETv0:Named Entities:Barack Obama': 29, + 'CSETv0:Named Entities:Delhi Metro Rail Corporation': 29, + 'CSETv0:Named Entities:Hyuandai Rotem': 29, + 'CSETv0:Named Entities:Jordan Peele': 29, + 'CSETv0:Named Entities:Tesla Autopilot': 29, + 'CSETv0:Named Entities:Tesla Model S': 29, + 'CSETv0:Named Entities:University of Washington': 29, + 'CSETv0:System Developer:FakeApp': 29, + 'CSETv0:System Developer:University of Washington': 29, + 'CSETv0:Technology Purveyor:Delhi Metro Rail Corporation': 29, + 'CSETv0:Technology Purveyor:Hyuandai Rotem': 29, + 'CSETv0:Technology Purveyor:Jordan Peele': 29, + 'CSETv0:Technology Purveyor:University of Washington': 29, + 'CSETv1:AI Task:content ranking': 29, + 'CSETv1:AI tools and methods:Sensor Data Processing': 29, + 'CSETv1:Data Inputs:Train position/status': 29, + 'CSETv1:Data Inputs:geospatial data': 29, + 'CSETv1:Data Inputs:track sensors': 29, + 'CSETv1:Infrastructure Sectors:communications': 29, + 'CSETv1:Operating Conditions:Testing': 29, + 'CSETv1:Physical System Type:Camera': 29, + 'CSETv1:Physical System Type:Metro train': 29, + 'CSETv1_Annotator-1:AI Task:autocomplete': 29, + 'CSETv1_Annotator-1:AI Task:object detection': 29, + 'CSETv1_Annotator-1:Annotator’s AI special interest intangible harm assessment:maybe': 29, + 'CSETv1_Annotator-1:Date of Incident Year:2011': 29, + 'CSETv1_Annotator-1:Location Region:--': 29, + 'CSETv1_Annotator-1:Physical System Type:Automobile': 29, + 'CSETv1_Annotator-1:Physical System Type:Metro train': 29, + 'CSETv1_Annotator-2:AI Task:Deepfake': 29, + 'CSETv1_Annotator-2:Data Inputs:Train position/status': 29, + 'CSETv1_Annotator-2:Data Inputs:track sensors': 29, + 'CSETv1_Annotator-2:Detrimental Content:maybe': 29, + 'CSETv1_Annotator-2:Operating Conditions:Testing': 29, + 'CSETv1_Annotator-2:Physical System Type:Metro train': 29, + 'CSETv1_Annotator-3:AI Task:autocomplete': 29, + 'CSETv1_Annotator-3:AI Task:content ranking': 29, + 'CSETv1_Annotator-3:Date of Incident Year:2011': 29, + 'GMF:Known AI Goal:Deepfake Video Generation': 29, + 'GMF:Known AI Goal:Hate Speech Detection': 29, + 'GMF:Known AI Technical Failure:Adversarial Data': 29, + 'GMF:Potential AI Technical Failure:Distributional Bias': 29, + 'GMF:Potential AI Technical Failure:Misuse': 29, + 'GMF:Potential AI Technology:3D reconstruction': 29, + 'GMF:Potential AI Technology:Other domain-specific approaches': 29, + 'CSETv0:AI Applications:Text generation': 28, + 'CSETv0:AI Applications:comprehension': 28, + 'CSETv0:AI Applications:computer vision': 28, + 'CSETv0:AI Applications:language output': 28, + 'CSETv0:AI Techniques:content creation': 28, + 'CSETv0:AI Techniques:driving lane detection': 28, + 'CSETv0:AI Techniques:image recognition': 28, + 'CSETv0:AI Techniques:language recognition natural language processing': 28, + 'CSETv0:AI Techniques:object detection': 28, + 'CSETv0:AI Techniques:scene segmentation': 28, + 'CSETv0:Location:Cambridge, MA': 28, + 'CSETv0:Location:Mountain View, CA': 28, + 'CSETv0:Location:Storey County, Nevada': 28, + 'CSETv0:Named Entities:Lexus': 28, + 'CSETv0:Named Entities:MIT Media Lab': 28, + 'CSETv0:Named Entities:Massachusetts Intstitute of Technology': 28, + 'CSETv0:Named Entities:Norman': 28, + 'CSETv0:Named Entities:Panasonic': 28, + 'CSETv0:Named Entities:Reddit': 28, + 'CSETv0:Named Entities:Santa Clara Valley Transportation Authority': 28, + 'CSETv0:Named Entities:Tay': 28, + 'CSETv0:Named Entities:Twitter': 28, + 'CSETv0:Named Entities:Xiaoice': 28, + 'CSETv0:System Developer:MIT Media Lab': 28, + 'CSETv0:Technology Purveyor:MIT Media Lab': 28, + 'CSETv0:Technology Purveyor:Twitter': 28, + 'CSETv1:AI Task:Generate Captions': 28, + 'CSETv1:AI Task:assembly': 28, + 'CSETv1:AI Task:image identification': 28, + 'CSETv1:AI Task:image interpretation': 28, + 'CSETv1:AI Task:production': 28, + 'CSETv1:AI tools and methods:prediction': 28, + 'CSETv1:Data Inputs:Captions': 28, + 'CSETv1:Data Inputs:image captions': 28, + 'CSETv1:Data Inputs:laser input': 28, + 'CSETv1:Data Inputs:video captions': 28, + 'CSETv1:Date of Incident Year:2009': 28, + 'CSETv1:Operating Conditions:Driverless car': 28, + 'CSETv1:Physical System Type:Vehicles (Lexus, Audi, Chrysler Pacifica)': 28, + 'CSETv1_Annotator-1:AI Task:assembly': 28, + 'CSETv1_Annotator-1:AI Task:image classification': 28, + 'CSETv1_Annotator-1:AI Task:image interpretation': 28, + 'CSETv1_Annotator-1:AI Task:production': 28, + 'CSETv1_Annotator-1:AI tools and methods:human language technology': 28, + 'CSETv1_Annotator-1:AI tools and methods:large language models': 28, + 'CSETv1_Annotator-1:Data Inputs:image captions': 28, + 'CSETv1_Annotator-1:Data Inputs:video captions': 28, + 'CSETv1_Annotator-1:Date of Incident Year:2024': 28, + 'CSETv1_Annotator-1:Physical System Type:Vehicles (Lexus, Audi, Chrysler Pacifica)': 28, + 'CSETv1_Annotator-2:AI Task:Generate Captions': 28, + 'CSETv1_Annotator-2:AI Task:Manufacturing Car': 28, + 'CSETv1_Annotator-2:Data Inputs:Captions': 28, + 'CSETv1_Annotator-2:Harm Distribution Basis:nation of origin, citizenship, immigrant status': 28, + 'CSETv1_Annotator-2:Harm Distribution Basis:religion': 28, + 'GMF:Known AI Technical Failure:Generalization Failure': 28, + 'GMF:Known AI Technical Failure:Limited Dataset': 28, + 'CSETv0:AI Applications:Early warning system': 27, + 'CSETv0:AI Applications:targeted advertising': 27, + 'CSETv0:AI Techniques:Google Adsense': 27, + 'CSETv0:AI Techniques:Oko satellites': 27, + 'CSETv0:Harm Type:Other:Reputational harm': 27, + 'CSETv0:Infrastructure Sectors:Critical manufacturing': 27, + 'CSETv0:Infrastructure Sectors:Nuclear': 27, + 'CSETv0:Location:Baunatal, Germany': 27, + 'CSETv0:Location:Soviet Union': 27, + 'CSETv0:Named Entities:Carnegie Mellon University': 27, + 'CSETv0:Named Entities:Harvard University': 27, + 'CSETv0:Named Entities:Harwin Cheng': 27, + 'CSETv0:Named Entities:Oko': 27, + 'CSETv0:Named Entities:Soviet Union': 27, + 'CSETv0:Named Entities:Tiffany Teng': 27, + 'CSETv0:Named Entities:United States': 27, + 'CSETv0:Named Entities:Volkswagen': 27, + 'CSETv0:Named Entities:William Santana Li': 27, + 'CSETv0:Named Entities:www.instantcheckmate.com': 27, + 'CSETv0:Physical System:Weapons system': 27, + 'CSETv0:System Developer:Soviet Union': 27, + 'CSETv0:Technology Purveyor:Instant Checkmate': 27, + 'CSETv0:Technology Purveyor:Soviet Union': 27, + 'CSETv0:Technology Purveyor:United States': 27, + 'CSETv0:Technology Purveyor:Volkswagen': 27, + 'CSETv1:AI Task:personalized online advertising': 27, + 'CSETv1:Data Inputs:satellite data': 27, + 'CSETv1:Data Inputs:sonar': 27, + 'CSETv1:Data Inputs:vibration detectors': 27, + 'CSETv1:Date of Incident Year:1983': 27, + 'CSETv1:Injuries:10': 27, + 'CSETv1_Annotator-1:AI Task:personalized online advertising': 27, + 'CSETv1_Annotator-1:Data Inputs:satellite data': 27, + 'CSETv1_Annotator-1:Date of Incident Year:1983': 27, + 'CSETv1_Annotator-1:Injuries:10': 27, + 'CSETv1_Annotator-1:Sector of Deployment:accommodation and food service activities': 27, + 'CSETv1_Annotator-2:AI Task:facial recognition': 27, + 'CSETv1_Annotator-2:AI tools and methods:Sensor Data Processing': 27, + 'CSETv1_Annotator-2:Data Inputs:sonar': 27, + 'CSETv1_Annotator-2:Data Inputs:vibration detectors': 27, + 'CSETv1_Annotator-2:Injuries:1': 27, + 'CSETv1_Annotator-2:Location Country (two letters):CN': 27, + "CSETv1_Annotator-2:Physical System Type:5' tall, 300 pound, egg shaped security robot": 27, + 'CSETv1_Annotator-3:Annotator’s AI special interest intangible harm assessment:maybe': 27, + 'CSETv1_Annotator-3:Clear link to Technology:maybe': 27, + 'CSETv1_Annotator-3:Date of Incident Year:1983': 27, + 'CSETv1_Annotator-3:Infrastructure Sectors:defense-industrial base': 27, + 'CSETv1_Annotator-3:Intentional Harm:Yes. Intentionally designed to perform harm but created an unintended harm (a different harm may have occurred)': 27, + 'CSETv1_Annotator-3:Location Country (two letters):RU': 27, + 'CSETv1_Annotator-3:Sector of Deployment:defense': 27, + 'CSETv1_Annotator-3:Tangible Harm:non-imminent risk of tangible harm (an issue) occurred': 27, + 'GMF:Known AI Goal:Threat Detection': 27, + 'GMF:Known AI Technical Failure:Black Swan Event': 27, + 'GMF:Known AI Technical Failure:Data or Labelling Noise': 27, + 'GMF:Known AI Technical Failure:Dataset Imbalance': 27, + 'GMF:Potential AI Technology:Face Detection': 27, + 'GMF:Potential AI Technology:Satellite Imaging': 27, + 'CSETv0:Location:Beitar Illit, Israel': 26, + 'CSETv0:Named Entities:Beitar Illit': 26, + 'CSETv0:Named Entities:Israel': 26, + 'CSETv0:Named Entities:Israeli Police': 26, + 'CSETv0:Sector of Deployment:Human health and social work activities': 26, + 'CSETv1:Data Inputs:Youtube videos': 26, + 'CSETv1:Data Inputs:traffic patterns': 26, + 'CSETv1:Location Country (two letters):PS': 26, + 'CSETv1:Location Country (two letters):VN': 26, + 'CSETv1_Annotator-1:AI Task:route optimization': 26, + 'CSETv1_Annotator-1:AI tools and methods:Dijkstra Algorithm': 26, + 'CSETv1_Annotator-1:AI tools and methods:shortest-path algorithm': 26, + 'CSETv1_Annotator-1:Data Inputs:GPS input': 26, + 'CSETv1_Annotator-1:Data Inputs:traffic patterns': 26, + 'CSETv1_Annotator-1:Data Inputs:user traffic reports': 26, + 'CSETv1_Annotator-1:Location Country (two letters):GB': 26, + 'CSETv1_Annotator-1:User Test in Controlled Conditions:yes': 26, + 'CSETv1_Annotator-2:AI Task:Translation': 26, + 'CSETv1_Annotator-2:Location Country (two letters):PS': 26, + 'CSETv1_Annotator-3:AI Task:Translation': 26, + 'CSETv1_Annotator-3:Location Country (two letters):PS': 26, + 'GMF:Known AI Goal:Translation': 26, + 'GMF:Potential AI Technical Failure:Misconfigured Aggregation': 26, + 'GMF:Potential AI Technology:Image Classification': 26, + 'GMF:Potential AI Technology:Intermediate modeling': 26, + 'GMF:Potential AI Technology:Multimodal Learning': 26, + 'CSETv0:Harm Type:Other': 25, + 'CSETv0:Location:Ningbo, China': 25, + 'CSETv0:Location:United States': 25, + 'CSETv0:Named Entities:Dong Mingzhu': 25, + 'CSETv0:Named Entities:Google Photos': 25, + 'CSETv0:Named Entities:Gree Electric Appliances': 25, + 'CSETv0:Named Entities:Ningbo': 25, + 'CSETv0:Physical System:Other:CCTV cameras, displays': 25, + 'CSETv0:System Developer:Ningbo traffic police': 25, + 'CSETv0:Technology Purveyor:Ningbo traffic police': 25, + 'CSETv1:AI Task:image classification': 25, + 'CSETv1:AI Task:jaywalking detection': 25, + 'CSETv1:Data Inputs:sex': 25, + 'CSETv1:Infrastructure Sectors:information technology': 25, + 'CSETv1:Lives Lost:2': 25, + 'CSETv1:Operating Conditions:night': 25, + 'CSETv1:Physical System Type:Tesla vehicle (Model 3, Model S, Model X)': 25, + 'CSETv1:User Test in Controlled Conditions:yes': 25, + 'CSETv1_Annotator-1:AI Task:jaywalking detection': 25, + 'CSETv1_Annotator-1:Data Inputs:camera footage': 25, + 'CSETv1_Annotator-1:Data Inputs:sex': 25, + 'CSETv1_Annotator-1:Date of Incident Year:2009': 25, + 'CSETv1_Annotator-1:Harm Distribution Basis:age': 25, + 'CSETv1_Annotator-1:Physical System Type:Tesla vehicle (Model 3, Model S, Model X)': 25, + 'CSETv1_Annotator-1:Physical System Type:autonomous car': 25, + 'CSETv1_Annotator-2:AI Task:Detect Jaywalkers': 25, + 'CSETv1_Annotator-2:Physical System Type:Camera': 25, + 'CSETv1_Annotator-3:AI Harm Level:AI tangible harm issue': 25, + 'GMF:Potential AI Technical Failure:Concept Drift': 25, + }, + 'CSETv1.AI System': { + yes: 1221, + no: 269, + maybe: 92, + }, + 'CSETv1.Lives Lost': { + '0': 1476, + '1': 76, + '2': 25, + '18': 2, + '51': 1, + '130': 3, + '144': 12, + '189': 19, + }, + 'GMF.Known AI Goal': { + 'Autonomous Driving': 45, + 'Face Recognition': 36, + 'AI Voice Assistant': 35, + 'Autonomous Drones': 31, + 'Automatic Stock Trading': 30, + 'Deepfake Video Generation': 29, + 'Hate Speech Detection': 29, + 'Threat Detection': 27, + Translation: 26, + 'Image Tagging': 24, + 'Content Recommendation': 14, + 'Content Search': 14, + 'Market Forecasting': 14, + 'NSFW Content Detection': 14, + Chatbot: 13, + 'Audio Localization': 11, + 'Gunshot Detection': 11, + 'Automatic Skill Assessment': 10, + Scheduling: 10, + 'Robotic Manipulation': 8, + 'Visual Art Generation': 7, + 'Smart Devices': 6, + 'Substance Detection': 6, + 'Code Generation': 5, + 'Image Cropping': 5, + 'Activity Tracking': 3, + 'Question Answering': 3, + 'Copyrighted Content Detection': 2, + 'Data Grouping': 2, + 'Social Media Content Generation': 2, + 'Automated Content Curation': 1, + 'Behavioral Modeling': 1, + 'Financial Processing': 1, + 'License Plate Recognition': 1, + 'Recidivism Prediction': 1, + }, + 'CSETv1.Data Inputs': { + text: 443, + images: 286, + video: 234, + 'sensor data': 147, + audio: 143, + lidar: 143, + radar: 131, + 'facial images': 100, + 'camera input': 95, + 'search queries': 68, + speech: 64, + 'video input': 61, + voice: 61, + photos: 58, + 'road data': 57, + 'spatial data': 52, + 'infrared camera': 51, + 'personal data': 51, + 'image alt tags': 47, + 'dot projector': 45, + GPS: 44, + 'income data': 39, + tabular: 39, + 'tax reports': 39, + 'Amazon account information': 37, + 'Facebook posts': 36, + 'air quality': 36, + 'Crime Records': 35, + Resume: 33, + traffic: 33, + age: 31, + 'personal information': 31, + 'radar input': 30, + 'stock data': 30, + 'stock orders': 30, + 'stock prices': 30, + 'thermal imaging': 30, + words: 30, + 'Train position/status': 29, + 'geospatial data': 29, + 'track sensors': 29, + Captions: 28, + 'image captions': 28, + 'laser input': 28, + 'video captions': 28, + 'satellite data': 27, + sonar: 27, + 'vibration detectors': 27, + 'Youtube videos': 26, + 'traffic patterns': 26, + sex: 25, + 'DAO token IDs': 24, + 'GNSS Antennae input': 24, + 'GPS input': 24, + code: 24, + 'inertial measurement unit input': 24, + 'odometer input': 24, + 'sales data': 24, + 'traffic light signals': 24, + 'user prompts': 24, + 'user traffic reports': 24, + 'previous criminal history': 23, + 'user queries': 23, + '137 questionnaire responses': 22, + 'Police reports': 22, + 'camera footage': 22, + 'criminal degree': 22, + 'route specifications': 22, + 'employee profile': 21, + 'infrared images': 21, + 'employee status': 20, + sensor: 17, + messages: 16, + 'criminal record': 15, + 'questionnaire responses': 15, + keywords: 14, + 'store camera footage': 14, + 'surveillance footage': 14, + 'census data': 13, + maps: 13, + names: 13, + 'news articles': 13, + 'past location of crime': 13, + 'past time of crime': 13, + 'past type of crime': 13, + 'patrol shifts': 13, + 'population density': 13, + 'game data': 11, + 'microphone inputs': 11, + 'patient data': 11, + schedules: 11, + 'security footage': 11, + selfies: 11, + 'store traffic': 10, + 'test scores': 10, + 'worker profiles': 10, + 'audio inputs': 9, + 'search terms': 9, + 'sensor input': 8, + 'treatment cost': 7, + 'Facebook accounts': 6, + 'Youtube content': 6, + 'credit report': 6, + income: 6, + motion: 6, + 'personal debt': 6, + 'personal finances': 6, + temperature: 6, + 'CDPH guidelines': 5, + 'Facebook ads': 5, + 'Instagram accounts': 5, + 'Uber accounts': 5, + department: 5, + 'driver behavior data': 5, + 'driver rating data': 5, + 'job role': 5, + 'medical personnel data': 5, + '148 million tax returns': 4, + '780000 audits': 4, + 'Uber driver account information': 4, + 'applicant data': 4, + bluetooth: 4, + cameras: 4, + 'census demographics': 4, + 'electronic health records': 4, + 'employee activity': 4, + 'employee performance data': 4, + 'geographical information': 4, + 'home valuations': 4, + 'homeowner-submitted details': 4, + 'house pictures': 4, + 'images from social media': 4, + 'images from websites': 4, + 'property records': 4, + 'real estate data': 4, + 'scraped images': 4, + 'social media accounts': 4, + 'student data': 4, + 'tax records': 4, + 'taxpayer names': 4, + 'taxpayer profile': 4, + 'user data': 4, + 'Employee Info': 3, + GPA: 3, + 'Product Info': 3, + RFID: 3, + 'Time off Task (TOT)': 3, + Unclear: 3, + 'Warehouse data': 3, + 'creatinine levels': 3, + 'email addresses': 3, + 'ethical judgments': 3, + 'ethical scenarios': 3, + 'medical data': 3, + race: 3, + 'student grades': 3, + usernames: 3, + 'voice input': 3, + 'water samples': 3, + 'worker performance data': 3, + 'Amazon products': 2, + 'Drake songs': 2, + 'Drake voice': 2, + 'The Weeknd songs': 2, + 'The Weeknd voice': 2, + 'TikTok posts': 2, + 'area of research interest': 2, + biometrics: 2, + 'body scans': 2, + 'faculty advisor': 2, + 'family relationships': 2, + 'letter of recommendation': 2, + 'letters of recommendation': 2, + 'license plate images': 2, + 'patient information': 2, + 'photos in Google photos': 2, + 'purchase history': 2, + 'reliability index': 2, + 'shift data': 2, + songs: 2, + 'statement of interest': 2, + 'university name': 2, + 'university previously attended': 2, + websites: 2, + '286 questions': 1, + 'Amazon items': 1, + 'Amazon purchases': 1, + 'Burmese text': 1, + 'Facebook comments': 1, + 'Facebook likes': 1, + 'Facebook reactions': 1, + 'Facebook reshares': 1, + 'Instagram posts': 1, + 'Internet usage': 1, + 'SAT and ACT scores': 1, + 'TikTok accounts': 1, + 'Twitter posts': 1, + acceleration: 1, + advertisements: 1, + 'alleged crimes': 1, + 'arrest photos': 1, + 'article text': 1, + 'audio recordings': 1, + 'biometric data': 1, + breaking: 1, + 'credits attempted': 1, + 'credits completed': 1, + 'dates of significance': 1, + 'delivery times': 1, + diagnosis: 1, + 'driver performance metrics': 1, + emails: 1, + 'estimated skills': 1, + 'health data': 1, + 'health measures': 1, + 'high school percentile': 1, + 'in-van camera input': 1, + 'income statements': 1, + ingredients: 1, + 'job application': 1, + 'living situation': 1, + 'marketing material': 1, + 'national IDs': 1, + 'phone calls': 1, + 'photo IDs': 1, + 'physical function': 1, + 'predicted final grades': 1, + 'property data': 1, + 'school attended': 1, + 'school principal assessments': 1, + "schools' historical IB results": 1, + 'seatbelt usage': 1, + 'social media posts': 1, + speed: 1, + 'standardized test results': 1, + 'statewide average test scores': 1, + 'student evaluations': 1, + 'student test results': 1, + symptoms: 1, + texting: 1, + 'uploaded images': 1, + 'user comments': 1, + 'user engagement': 1, + 'user reports': 1, + 'voice recordings': 1, + 'webpage content': 1, + 'website code': 1, + 'welfare recipient data': 1, + 'x-ray images': 1, + 'youtube history': 1, + 'zip code': 1, + }, + is_incident_report: { + true: 3871, + }, + epoch_incident_date: { + '433382400': 27, + '706752000': 2, + '828489600': 1, + '889056000': 4, + '921542400': 3, + '1048291200': 1, + '1049241600': 4, + '1054425600': 3, + '1057017600': 2, + '1140825600': 3, + '1185408000': 7, + '1211500800': 24, + '1214870400': 1, + '1216339200': 4, + '1238371200': 1, + '1251763200': 1, + '1273276800': 30, + '1301961600': 29, + '1303084800': 1, + '1316476800': 2, + '1325376000': 1, + '1325721600': 1, + '1330128000': 7, + '1336089600': 4, + '1343779200': 10, + '1349740800': 11, + '1354320000': 2, + '1355097600': 2, + '1358899200': 27, + '1375142400': 2, + '1377993600': 5, + '1379030400': 3, + '1380585600': 12, + '1385510400': 1, + '1388534400': 1, + '1390262400': 6, + '1393891200': 5, + '1397606400': 2, + '1399507200': 2, + '1405382400': 27, + '1407974400': 10, + '1408060800': 10, + '1409011200': 2, + '1412121600': 2, + '1413590400': 20, + '1418083200': 3, + '1419206400': 1, + '1425168000': 16, + '1427155200': 1, + '1427846400': 3, + '1428105600': 11, + '1430438400': 2, + '1431302400': 11, + '1431993600': 14, + '1433116800': 2, + '1433289600': 24, + '1434240000': 1, + '1435708800': 39, + '1435795200': 12, + '1436745600': 12, + '1436918400': 1, + '1438387200': 1, + '1441065600': 13, + '1443139200': 5, + '1446508800': 22, + '1447372800': 3, + '1447804800': 13, + '1449273600': 35, + '1451606400': 1, + '1453248000': 4, + '1454284800': 1, + '1454457600': 1, + '1455062400': 7, + '1456790400': 8, + '1458777600': 28, + '1459382400': 18, + '1459468800': 4, + '1459814400': 4, + '1460160000': 1, + '1460678400': 1, + '1463961600': 37, + '1464220800': 3, + '1464739200': 10, + '1464825600': 12, + '1465948800': 5, + '1466121600': 24, + '1467244800': 22, + '1467331200': 29, + '1468281600': 27, + '1469059200': 1, + '1470787200': 33, + '1472688000': 5, + '1473033600': 10, + '1473120000': 9, + '1473379200': 13, + '1474848000': 28, + '1475280000': 2, + '1475884800': 29, + '1479254400': 2, + '1481068800': 22, + '1481760000': 1, + '1482364800': 1, + '1483056000': 16, + '1483142400': 4, + '1483228800': 1, + '1484438400': 4, + '1485907200': 1, + '1487894400': 6, + '1488153600': 9, + '1488326400': 2, + '1489363200': 3, + '1489536000': 2, + '1491436800': 5, + '1491523200': 2, + '1491782400': 3, + '1492041600': 10, + '1492214400': 1, + '1493078400': 23, + '1493337600': 1, + '1493596800': 1, + '1494201600': 1, + '1495152000': 3, + '1496448000': 8, + '1497484800': 10, + '1498867200': 35, + '1499040000': 2, + '1499644800': 7, + '1500249600': 30, + '1501545600': 2, + '1501632000': 16, + '1502755200': 2, + '1503705600': 4, + '1504742400': 1, + '1505260800': 45, + '1505433600': 1, + '1505692800': 1, + '1505952000': 2, + '1506470400': 1, + '1507593600': 5, + '1507766400': 5, + '1508198400': 26, + '1508976000': 7, + '1509494400': 1, + '1510099200': 24, + '1510185600': 4, + '1510704000': 8, + '1511222400': 1, + '1512259200': 29, + '1512518400': 22, + '1515715200': 3, + '1515974400': 2, + '1516579200': 5, + '1516838400': 1, + '1517443200': 1, + '1518652800': 2, + '1521331200': 25, + '1521763200': 14, + '1522627200': 28, + '1523404800': 4, + '1523491200': 3, + '1524182400': 2, + '1524441600': 2, + '1524960000': 4, + '1525132800': 6, + '1525910400': 4, + '1525996800': 1, + '1527292800': 4, + '1527552000': 4, + '1527811200': 1, + '1528675200': 3, + '1530489600': 1, + '1530662400': 1, + '1530921600': 4, + '1532044800': 2, + '1532131200': 3, + '1532563200': 1, + '1533081600': 3, + '1534118400': 4, + '1534291200': 5, + '1535760000': 6, + '1538352000': 2, + '1540598400': 19, + '1541462400': 25, + '1541635200': 5, + '1541721600': 2, + '1542931200': 2, + '1543363200': 2, + '1543622400': 25, + '1543795200': 1, + '1543968000': 17, + '1546300800': 1, + '1548806400': 4, + '1548979200': 3, + '1549238400': 3, + '1550102400': 3, + '1551398400': 6, + '1552262400': 3, + '1552608000': 1, + '1554076800': 7, + '1554249600': 3, + '1559260800': 3, + '1559347200': 5, + '1559520000': 2, + '1561161600': 8, + '1561420800': 2, + '1561939200': 3, + '1562112000': 3, + '1562371200': 3, + '1562716800': 1, + '1562803200': 2, + '1564531200': 3, + '1566604800': 1, + '1567728000': 2, + '1569715200': 3, + '1570147200': 5, + '1570492800': 3, + '1570579200': 1, + '1571097600': 6, + '1571875200': 7, + '1572998400': 4, + '1573430400': 6, + '1573516800': 6, + '1573776000': 4, + '1574985600': 4, + '1575676800': 2, + '1577577600': 8, + '1577836800': 6, + '1577923200': 1, + '1579046400': 4, + '1579305600': 1, + '1579564800': 6, + '1580342400': 11, + '1580947200': 1, + '1581033600': 2, + '1582070400': 3, + '1582156800': 3, + '1582502400': 4, + '1583712000': 4, + '1584316800': 4, + '1584403200': 1, + '1584921600': 2, + '1585267200': 4, + '1585699200': 1, + '1586822400': 2, + '1586908800': 1, + '1588032000': 3, + '1588377600': 2, + '1590883200': 9, + '1590969600': 4, + '1591401600': 12, + '1591574400': 2, + '1591747200': 1, + '1591833600': 2, + '1592006400': 2, + '1592179200': 2, + '1592352000': 1, + '1592611200': 8, + '1592956800': 4, + '1593216000': 1, + '1593302400': 6, + '1593561600': 9, + '1593993600': 1, + '1595030400': 1, + '1595894400': 8, + '1596412800': 1, + '1596585600': 2, + '1596672000': 3, + '1596758400': 3, + '1597276800': 8, + '1598918400': 1, + '1600128000': 2, + '1600387200': 5, + '1600560000': 1, + '1601337600': 3, + '1601683200': 3, + '1601942400': 2, + '1602028800': 1, + '1602115200': 2, + '1602201600': 2, + '1602720000': 1, + '1603152000': 5, + '1603238400': 2, + '1603324800': 1, + '1603497600': 2, + '1603929600': 1, + '1604188800': 3, + '1605052800': 4, + '1605398400': 1, + '1606435200': 2, + '1607040000': 2, + '1607212800': 1, + '1607990400': 2, + '1608249600': 5, + '1608681600': 13, + '1608768000': 1, + '1608854400': 2, + '1609027200': 1, + '1609459200': 4, + '1610236800': 4, + '1610323200': 1, + '1610496000': 2, + '1610668800': 1, + '1611187200': 2, + '1612137600': 1, + '1612483200': 3, + '1613001600': 1, + '1613088000': 1, + '1613433600': 1, + '1614124800': 1, + '1614297600': 2, + '1614556800': 1, + '1614643200': 4, + '1615766400': 2, + '1615939200': 7, + '1617235200': 3, + '1618272000': 1, + '1618617600': 9, + '1618963200': 1, + '1619568000': 1, + '1619654400': 2, + '1619827200': 8, + '1620259200': 3, + '1621296000': 7, + '1621728000': 1, + '1622160000': 6, + '1622592000': 7, + '1622678400': 4, + '1623456000': 2, + '1624924800': 5, + '1625097600': 1, + '1625184000': 1, + '1625356800': 4, + '1625875200': 3, + '1625961600': 2, + '1626134400': 2, + '1626393600': 4, + '1626739200': 1, + '1626998400': 3, + '1627603200': 1, + '1627776000': 14, + '1627948800': 4, + '1628985600': 4, + '1630454400': 3, + '1631836800': 3, + '1632096000': 2, + '1633046400': 4, + '1633478400': 1, + '1634256000': 3, + '1634515200': 1, + '1634774400': 1, + '1634860800': 3, + '1635379200': 10, + '1635724800': 8, + '1635811200': 4, + '1635897600': 1, + '1636243200': 1, + '1636502400': 5, + '1637452800': 3, + '1638748800': 18, + '1638921600': 1, + '1639180800': 7, + '1640390400': 2, + '1640476800': 2, + '1640563200': 2, + '1640995200': 1, + '1642118400': 1, + '1642204800': 10, + '1642809600': 3, + '1643155200': 1, + '1643932800': 5, + '1644451200': 3, + '1644537600': 6, + '1644883200': 1, + '1645574400': 1, + '1645747200': 4, + '1646006400': 4, + '1646092800': 7, + '1646179200': 1, + '1646265600': 1, + '1646611200': 2, + '1647302400': 3, + '1647388800': 9, + '1647475200': 2, + '1647907200': 2, + '1648771200': 8, + '1648944000': 1, + '1649116800': 2, + '1649203200': 2, + '1649808000': 1, + '1650326400': 5, + '1650412800': 3, + '1650499200': 8, + '1651968000': 2, + '1652400000': 2, + '1652659200': 17, + '1652832000': 4, + '1653436800': 3, + '1654041600': 1, + '1654214400': 12, + '1654905600': 4, + '1654992000': 2, + '1656374400': 1, + '1658102400': 1, + '1658361600': 5, + '1658448000': 3, + '1658620800': 3, + '1658793600': 2, + '1659830400': 3, + '1660521600': 3, + '1660694400': 1, + '1661040000': 2, + '1661126400': 3, + '1661385600': 2, + '1661731200': 1, + '1662508800': 2, + '1662854400': 2, + '1663027200': 3, + '1663200000': 18, + '1663891200': 1, + '1664755200': 2, + '1664841600': 8, + '1665446400': 1, + '1665878400': 5, + '1667952000': 1, + '1668124800': 14, + '1668470400': 6, + '1668902400': 9, + '1669075200': 6, + '1669248000': 8, + '1669334400': 9, + '1669507200': 1, + '1669766400': 18, + '1669852800': 15, + '1671408000': 23, + '1671580800': 26, + '1672185600': 10, + '1672531200': 5, + '1672704000': 8, + '1672876800': 1, + '1673395200': 9, + '1674000000': 4, + '1674172800': 3, + '1674259200': 2, + '1675036800': 19, + '1675296000': 2, + '1675382400': 6, + '1675641600': 5, + '1675728000': 18, + '1675814400': 4, + '1675987200': 2, + '1676160000': 11, + '1676332800': 12, + '1676419200': 3, + '1676505600': 23, + '1676851200': 4, + '1677024000': 1, + '1677456000': 2, + '1677542400': 22, + '1677801600': 2, + '1677974400': 5, + '1678492800': 2, + '1678838400': 5, + '1679011200': 2, + '1679270400': 2, + '1679356800': 11, + '1679529600': 2, + '1679616000': 5, + '1679875200': 7, + '1680048000': 2, + '1680220800': 5, + '1680307200': 3, + '1680912000': 2, + '1681430400': 7, + '1681689600': 2, + '1682380800': 1, + '1682899200': 1, + '1683072000': 2, + '1683158400': 58, + '1683763200': 22, + '1683849600': 1, + '1684108800': 10, + '1684368000': 1, + '1684627200': 1, + '1684713600': 19, + '1684886400': 2, + '1685232000': 1, + '1685318400': 46, + '1685577600': 3, + '1685923200': 2, + '1686096000': 1, + '1686528000': 1, + '1686614400': 4, + '1687392000': 2, + '1687478400': 2, + '1687564800': 1, + '1687737600': 1, + '1687910400': 1, + '1688083200': 3, + '1688342400': 1, + '1688515200': 1, + '1689033600': 2, + '1689552000': 1, + '1689897600': 1, + '1690156800': 3, + '1691452800': 4, + '1691625600': 1, + '1691712000': 1, + '1691971200': 2, + '1692144000': 2, + '1693094400': 1, + '1693353600': 1, + '1693612800': 1, + '1694390400': 1, + '1694908800': 10, + '1695081600': 1, + '1696204800': 27, + '1696377600': 1, + '1696636800': 14, + '1696723200': 10, + '1697414400': 44, + '1697500800': 1, + '1698105600': 1, + '1698278400': 3, + '1698710400': 17, + '1698883200': 1, + '1699401600': 2, + '1699488000': 3, + '1699574400': 3, + '1700697600': 1, + '1701043200': 43, + '1701388800': 3, + '1701820800': 3, + '1702252800': 8, + '1702339200': 12, + '1702512000': 3, + '1702684800': 1, + '1702857600': 6, + '1703030400': 32, + '1703548800': 30, + '1703635200': 1, + '1703808000': 1, + '1704758400': 8, + '1704931200': 1, + '1705017600': 5, + '1705104000': 1, + '1705276800': 4, + '1705536000': 2, + '1705795200': 14, + '1705968000': 2, + '1706054400': 31, + '1706400000': 17, + '1706572800': 2, + '1706745600': 3, + '1706832000': 21, + '1707091200': 1, + '1707264000': 2, + '1707350400': 1, + '1707609600': 1, + '1707696000': 1, + '1707782400': 6, + '1707868800': 7, + '1708128000': 1, + '1708214400': 6, + '1708300800': 1, + '1708387200': 17, + '1708473600': 35, + '1708560000': 1, + '1709078400': 34, + '1709251200': 6, + '1709510400': 1, + '1709683200': 3, + '1710374400': 7, + '1710979200': 1, + '1711065600': 3, + '1711152000': 3, + '1711411200': 9, + '1711584000': 2, + '1711670400': 2, + '1711929600': 3, + '1712016000': 3, + '1712102400': 9, + '1712188800': 2, + '1712275200': 1, + '1712361600': 3, + '1712448000': 2, + '1712534400': 1, + '1713052800': 1, + '1713139200': 1, + '1713225600': 2, + '1713484800': 10, + '1713830400': 1, + '1713916800': 3, + '1714003200': 2, + '1714089600': 4, + '1714348800': 2, + '1714694400': 3, + '1715558400': 2, + '1715644800': 9, + '1715731200': 2, + '1715817600': 1, + '1716163200': 16, + '1716249600': 1, + '1716422400': 4, + '1716508800': 16, + '1716595200': 2, + '1716681600': 1, + '1716940800': 1, + '1717027200': 2, + '1717113600': 2, + '1717200000': 3, + '1717459200': 1, + '1717632000': 1, + '1717718400': 5, + '1717804800': 2, + '1717891200': 2, + '1718150400': 2, + '1718236800': 1, + '1718668800': 3, + '1718755200': 1, + '1718841600': 1, + '1718928000': 2, + '1719014400': 1, + '1719100800': 5, + '1719273600': 1, + '1719360000': 4, + '1719446400': 2, + '1719619200': 1, + '1719792000': 4, + '1719878400': 2, + '1719964800': 2, + '1720051200': 2, + '1720310400': 2, + '1720569600': 1, + '1720828800': 1, + '1721088000': 1, + '1721260800': 1, + '1721520000': 2, + '1721606400': 1, + '1721692800': 1, + '1721865600': 2, + '1721952000': 3, + '1722211200': 1, + '1722297600': 5, + '1722902400': 1, + '1723075200': 1, + '1723507200': 3, + '1723593600': 4, + '1723766400': 3, + '1723939200': 4, + '1724112000': 1, + '1724198400': 1, + '1724371200': 1, + '1724803200': 7, + '1725235200': 2, + '1725321600': 2, + '1725408000': 3, + '1725753600': 1, + '1725840000': 2, + '1726185600': 2, + '1726617600': 1, + '1726704000': 12, + '1727136000': 8, + '1727222400': 7, + '1727481600': 2, + '1727740800': 3, + '1727827200': 13, + '1728172800': 1, + '1728259200': 20, + '1728345600': 2, + '1728432000': 1, + '1728604800': 1, + '1728777600': 3, + '1728950400': 7, + '1729036800': 11, + '1729209600': 2, + '1729468800': 2, + '1729641600': 1, + '1729728000': 1, + '1729900800': 1, + '1730505600': 1, + '1731456000': 1, + '1732060800': 2, + }, + 'CSETv1.AI Harm Level': { + none: 1003, + 'AI tangible harm event': 373, + 'AI tangible harm near-miss': 104, + 'AI tangible harm issue': 66, + unclear: 36, + }, + 'CSETv1.Tangible Harm': { + 'tangible harm definitively occurred': 678, + 'no tangible harm, near-miss, or issue': 618, + 'imminent risk of tangible harm (near miss) did occur': 180, + 'non-imminent risk of tangible harm (an issue) occurred': 97, + unclear: 9, + }, + epoch_date_published: { + '1062028800': 1, + '1084579200': 1, + '1085011200': 1, + '1095292800': 1, + '1102723200': 1, + '1170806400': 1, + '1211500800': 1, + '1214870400': 1, + '1219104000': 1, + '1238803200': 1, + '1239494400': 5, + '1239580800': 7, + '1239667200': 5, + '1239840000': 1, + '1240012800': 1, + '1240531200': 1, + '1242259200': 1, + '1273276800': 1, + '1282176000': 1, + '1285804800': 1, + '1301961600': 2, + '1302048000': 1, + '1303430400': 1, + '1303862400': 1, + '1304553600': 1, + '1308009600': 1, + '1316476800': 1, + '1325721600': 2, + '1329350400': 2, + '1330128000': 1, + '1330300800': 1, + '1330387200': 1, + '1330560000': 1, + '1332720000': 1, + '1332892800': 1, + '1347753600': 1, + '1348099200': 1, + '1348704000': 1, + '1351641600': 1, + '1352160000': 1, + '1357948800': 1, + '1359417600': 1, + '1359676800': 1, + '1359936000': 5, + '1360022400': 5, + '1360108800': 4, + '1360195200': 2, + '1360800000': 1, + '1364860800': 1, + '1366070400': 1, + '1366243200': 1, + '1368403200': 1, + '1368489600': 1, + '1369785600': 1, + '1375142400': 1, + '1375747200': 1, + '1376697600': 1, + '1380153600': 1, + '1382140800': 1, + '1382572800': 1, + '1390262400': 1, + '1396569600': 1, + '1396656000': 1, + '1398211200': 1, + '1398643200': 1, + '1399334400': 1, + '1400630400': 1, + '1400803200': 2, + '1405382400': 1, + '1407283200': 2, + '1407369600': 1, + '1407974400': 1, + '1411689600': 1, + '1413590400': 1, + '1415318400': 1, + '1418083200': 1, + '1419379200': 1, + '1419811200': 2, + '1423267200': 1, + '1427932800': 1, + '1428105600': 1, + '1428537600': 2, + '1428624000': 2, + '1428710400': 1, + '1429056000': 1, + '1429142400': 1, + '1429574400': 2, + '1429660800': 7, + '1429747200': 5, + '1430092800': 1, + '1430784000': 1, + '1430870400': 1, + '1431302400': 1, + '1431993600': 1, + '1433289600': 1, + '1433376000': 1, + '1435276800': 3, + '1435363200': 1, + '1435449600': 1, + '1435536000': 1, + '1435622400': 1, + '1435708800': 9, + '1435795200': 28, + '1435881600': 1, + '1435968000': 1, + '1436140800': 2, + '1436313600': 1, + '1436572800': 1, + '1436745600': 2, + '1437350400': 1, + '1437436800': 6, + '1437523200': 2, + '1437868800': 1, + '1438560000': 1, + '1439078400': 1, + '1439424000': 5, + '1439510400': 2, + '1439683200': 1, + '1440633600': 1, + '1441065600': 1, + '1443052800': 2, + '1443571200': 1, + '1446422400': 1, + '1446508800': 6, + '1446595200': 1, + '1446681600': 1, + '1446854400': 1, + '1447286400': 1, + '1447804800': 1, + '1448409600': 1, + '1449273600': 1, + '1451606400': 1, + '1452729600': 1, + '1454112000': 1, + '1454457600': 1, + '1454544000': 1, + '1455062400': 2, + '1455235200': 1, + '1455840000': 1, + '1456444800': 1, + '1456704000': 3, + '1456790400': 9, + '1457049600': 1, + '1457568000': 5, + '1458000000': 1, + '1458691200': 1, + '1458777600': 6, + '1458864000': 7, + '1458950400': 3, + '1459036800': 2, + '1459296000': 1, + '1459382400': 1, + '1459555200': 1, + '1459987200': 2, + '1460073600': 2, + '1460419200': 1, + '1461283200': 1, + '1462060800': 1, + '1463961600': 6, + '1464048000': 1, + '1464220800': 3, + '1464393600': 1, + '1464825600': 2, + '1464912000': 3, + '1464998400': 2, + '1465084800': 1, + '1465171200': 1, + '1465257600': 2, + '1465344000': 2, + '1465430400': 4, + '1465516800': 7, + '1466121600': 1, + '1466208000': 3, + '1466294400': 1, + '1466380800': 1, + '1466553600': 1, + '1466812800': 2, + '1466899200': 1, + '1467244800': 1, + '1467331200': 13, + '1467417600': 7, + '1467504000': 1, + '1467590400': 1, + '1467676800': 1, + '1467849600': 1, + '1467936000': 1, + '1468195200': 1, + '1468281600': 4, + '1468368000': 12, + '1468454400': 6, + '1468540800': 3, + '1468800000': 1, + '1469059200': 1, + '1469232000': 1, + '1469491200': 1, + '1470441600': 2, + '1470700800': 1, + '1470787200': 2, + '1470873600': 1, + '1471392000': 1, + '1471564800': 1, + '1471824000': 2, + '1471996800': 1, + '1472083200': 1, + '1472428800': 1, + '1472601600': 1, + '1473033600': 1, + '1473120000': 2, + '1473206400': 1, + '1473292800': 4, + '1473379200': 5, + '1473465600': 1, + '1473811200': 2, + '1474243200': 1, + '1474329600': 1, + '1474675200': 1, + '1474761600': 2, + '1474848000': 5, + '1475798400': 1, + '1475884800': 1, + '1476057600': 1, + '1476144000': 1, + '1477267200': 1, + '1477872000': 1, + '1479427200': 4, + '1480550400': 1, + '1480636800': 1, + '1480723200': 1, + '1480982400': 1, + '1481068800': 10, + '1481155200': 9, + '1481760000': 1, + '1481846400': 2, + '1482019200': 1, + '1482278400': 1, + '1482364800': 1, + '1482537600': 1, + '1482883200': 3, + '1483056000': 1, + '1483142400': 4, + '1483315200': 1, + '1483401600': 2, + '1483488000': 5, + '1483574400': 3, + '1483660800': 7, + '1483747200': 2, + '1483833600': 2, + '1483920000': 10, + '1484006400': 4, + '1484092800': 3, + '1484524800': 3, + '1484611200': 2, + '1485216000': 1, + '1485561600': 1, + '1486339200': 1, + '1486425600': 2, + '1486857600': 1, + '1487894400': 2, + '1487980800': 2, + '1488067200': 1, + '1488153600': 3, + '1488240000': 1, + '1488326400': 5, + '1488499200': 2, + '1488585600': 1, + '1488931200': 2, + '1489622400': 1, + '1489795200': 1, + '1490227200': 3, + '1490572800': 1, + '1491264000': 1, + '1491782400': 4, + '1491868800': 1, + '1492041600': 1, + '1492128000': 1, + '1492473600': 1, + '1492560000': 1, + '1493078400': 6, + '1493164800': 1, + '1493251200': 1, + '1493424000': 1, + '1493596800': 1, + '1493769600': 1, + '1493856000': 1, + '1494201600': 1, + '1495152000': 2, + '1495238400': 2, + '1496102400': 1, + '1497225600': 1, + '1497312000': 2, + '1497830400': 1, + '1498003200': 1, + '1498089600': 2, + '1498694400': 1, + '1499040000': 1, + '1499299200': 1, + '1499644800': 3, + '1499731200': 5, + '1499817600': 2, + '1499904000': 2, + '1499990400': 3, + '1500249600': 2, + '1500336000': 21, + '1500422400': 10, + '1500595200': 2, + '1500854400': 1, + '1501027200': 1, + '1501200000': 1, + '1501545600': 1, + '1501718400': 7, + '1501804800': 5, + '1501891200': 3, + '1502064000': 1, + '1502150400': 1, + '1502236800': 2, + '1502323200': 11, + '1502496000': 1, + '1502755200': 1, + '1502928000': 1, + '1503014400': 2, + '1503187200': 1, + '1503273600': 2, + '1503360000': 1, + '1503446400': 1, + '1504224000': 1, + '1504742400': 2, + '1504915200': 1, + '1505088000': 1, + '1505260800': 4, + '1505347200': 4, + '1505433600': 1, + '1505692800': 6, + '1505779200': 3, + '1505952000': 2, + '1506038400': 2, + '1506470400': 1, + '1506556800': 1, + '1506729600': 1, + '1507248000': 1, + '1507507200': 1, + '1507680000': 3, + '1507766400': 3, + '1507852800': 2, + '1507939200': 1, + '1508198400': 1, + '1508630400': 2, + '1508716800': 4, + '1508803200': 12, + '1508889600': 4, + '1508976000': 6, + '1509062400': 3, + '1509235200': 2, + '1509321600': 1, + '1509408000': 5, + '1509494400': 3, + '1509667200': 1, + '1509840000': 1, + '1509926400': 2, + '1510012800': 1, + '1510099200': 2, + '1510185600': 14, + '1510272000': 7, + '1510358400': 2, + '1510531200': 8, + '1510617600': 6, + '1510704000': 4, + '1510876800': 1, + '1510963200': 1, + '1511308800': 1, + '1511740800': 1, + '1511827200': 3, + '1512000000': 1, + '1512086400': 1, + '1512259200': 1, + '1512518400': 2, + '1512604800': 4, + '1512691200': 11, + '1512777600': 4, + '1513036800': 2, + '1513123200': 4, + '1513209600': 1, + '1513296000': 2, + '1513382400': 1, + '1513468800': 1, + '1513555200': 1, + '1513641600': 5, + '1513728000': 11, + '1513814400': 7, + '1514764800': 2, + '1515628800': 3, + '1515715200': 2, + '1516060800': 2, + '1516147200': 7, + '1516233600': 6, + '1516579200': 3, + '1516752000': 4, + '1516838400': 2, + '1517529600': 2, + '1517788800': 3, + '1518566400': 3, + '1519776000': 2, + '1521417600': 2, + '1521504000': 7, + '1521763200': 3, + '1522281600': 6, + '1522454400': 6, + '1522627200': 5, + '1522800000': 2, + '1523404800': 5, + '1523491200': 3, + '1523577600': 4, + '1523664000': 3, + '1523836800': 4, + '1523923200': 3, + '1524009600': 5, + '1524096000': 2, + '1524182400': 2, + '1524528000': 2, + '1525132800': 3, + '1525478400': 2, + '1525651200': 4, + '1525737600': 3, + '1526342400': 2, + '1526428800': 2, + '1526515200': 3, + '1527120000': 2, + '1527206400': 2, + '1527465600': 2, + '1527552000': 4, + '1528070400': 3, + '1528156800': 2, + '1528243200': 2, + '1528329600': 11, + '1528416000': 5, + '1528761600': 2, + '1528848000': 5, + '1529020800': 2, + '1529539200': 5, + '1529625600': 8, + '1529712000': 3, + '1529884800': 6, + '1531180800': 2, + '1532390400': 2, + '1532476800': 2, + '1533686400': 2, + '1536192000': 2, + '1536624000': 2, + '1537228800': 2, + '1537401600': 3, + '1537488000': 3, + '1537833600': 2, + '1537920000': 4, + '1538092800': 2, + '1538438400': 2, + '1539129600': 7, + '1539216000': 14, + '1539302400': 4, + '1540857600': 2, + '1541635200': 5, + '1541808000': 2, + '1542326400': 2, + '1542844800': 2, + '1542931200': 8, + '1543017600': 2, + '1543190400': 2, + '1543276800': 6, + '1543363200': 6, + '1543449600': 4, + '1543622400': 3, + '1543708800': 3, + '1543795200': 6, + '1543881600': 3, + '1543968000': 5, + '1544054400': 11, + '1544140800': 6, + '1544486400': 2, + '1547510400': 2, + '1549324800': 2, + '1553126400': 2, + '1553731200': 2, + '1555113600': 4, + '1555200000': 7, + '1555977600': 3, + '1557360000': 2, + '1557964800': 2, + '1561420800': 2, + '1568592000': 2, + '1570147200': 3, + '1570665600': 2, + '1571270400': 3, + '1571875200': 3, + '1572998400': 4, + '1573516800': 4, + '1574121600': 2, + '1579305600': 3, + '1580256000': 2, + '1581292800': 2, + '1582070400': 2, + '1582588800': 2, + '1582675200': 3, + '1582761600': 3, + '1588291200': 2, + '1588896000': 2, + '1590105600': 2, + '1590710400': 3, + '1590796800': 4, + '1590969600': 5, + '1591056000': 2, + '1591660800': 3, + '1591747200': 2, + '1592956800': 3, + '1594339200': 2, + '1595203200': 2, + '1595980800': 2, + '1596067200': 3, + '1596153600': 2, + '1596499200': 5, + '1596585600': 3, + '1597104000': 2, + '1597190400': 2, + '1597622400': 3, + '1597795200': 2, + '1597881600': 2, + '1597968000': 2, + '1598486400': 2, + '1599091200': 2, + '1599609600': 2, + '1599696000': 2, + '1599782400': 2, + '1600646400': 2, + '1601337600': 2, + '1601942400': 3, + '1602028800': 2, + '1602201600': 4, + '1602720000': 2, + '1603152000': 2, + '1603324800': 2, + '1603670400': 3, + '1605744000': 2, + '1605830400': 2, + '1606262400': 2, + '1606521600': 3, + '1606780800': 2, + '1608249600': 4, + '1608422400': 2, + '1608768000': 2, + '1609027200': 3, + '1609286400': 2, + '1609545600': 2, + '1609718400': 2, + '1609804800': 2, + '1610236800': 2, + '1610323200': 2, + '1610409600': 3, + '1610496000': 4, + '1610582400': 3, + '1610928000': 2, + '1611014400': 2, + '1611878400': 4, + '1612137600': 2, + '1612310400': 2, + '1612569600': 2, + '1612828800': 2, + '1613001600': 2, + '1613433600': 2, + '1613692800': 2, + '1614211200': 2, + '1614556800': 2, + '1614643200': 2, + '1615334400': 2, + '1615507200': 3, + '1615939200': 3, + '1616025600': 3, + '1616112000': 2, + '1617321600': 2, + '1617926400': 3, + '1618704000': 3, + '1618790400': 2, + '1618876800': 2, + '1619395200': 2, + '1619568000': 3, + '1621382400': 2, + '1621468800': 2, + '1622592000': 3, + '1622678400': 3, + '1622764800': 3, + '1624838400': 2, + '1624924800': 4, + '1625184000': 2, + '1625443200': 3, + '1626220800': 3, + '1626652800': 4, + '1626825600': 2, + '1627257600': 2, + '1627603200': 2, + '1628208000': 2, + '1628467200': 2, + '1629331200': 2, + '1629849600': 2, + '1632096000': 2, + '1634083200': 4, + '1634601600': 2, + '1634688000': 3, + '1634860800': 3, + '1635811200': 2, + '1635897600': 3, + '1636329600': 4, + '1636416000': 3, + '1636502400': 2, + '1636588800': 3, + '1636761600': 2, + '1636934400': 6, + '1637193600': 2, + '1637452800': 3, + '1638144000': 3, + '1638403200': 2, + '1638662400': 2, + '1639008000': 4, + '1639094400': 2, + '1639353600': 2, + '1639440000': 8, + '1639526400': 4, + '1639612800': 4, + '1640649600': 2, + '1641427200': 2, + '1641772800': 3, + '1642464000': 3, + '1643241600': 3, + '1643760000': 4, + '1643932800': 2, + '1644192000': 2, + '1644278400': 2, + '1644796800': 2, + '1645574400': 2, + '1646006400': 3, + '1646092800': 2, + '1646611200': 3, + '1646697600': 5, + '1646784000': 3, + '1647388800': 5, + '1647475200': 7, + '1647820800': 3, + '1647907200': 3, + '1647993600': 2, + '1648166400': 3, + '1648425600': 2, + '1648512000': 3, + '1648684800': 2, + '1648771200': 2, + '1649030400': 2, + '1649116800': 2, + '1649548800': 4, + '1649635200': 2, + '1650585600': 3, + '1650672000': 3, + '1650844800': 4, + '1650931200': 2, + '1653177600': 2, + '1653609600': 2, + '1653868800': 2, + '1654128000': 2, + '1654214400': 2, + '1654473600': 4, + '1655164800': 2, + '1655251200': 2, + '1657065600': 2, + '1657152000': 3, + '1657238400': 3, + '1657497600': 2, + '1658102400': 2, + '1658361600': 2, + '1658448000': 5, + '1658620800': 4, + '1658707200': 3, + '1658793600': 3, + '1658880000': 2, + '1658966400': 3, + '1659657600': 2, + '1659744000': 3, + '1659916800': 2, + '1660089600': 2, + '1660521600': 2, + '1661126400': 4, + '1661212800': 2, + '1661299200': 2, + '1661385600': 2, + '1661472000': 2, + '1661904000': 2, + '1661990400': 2, + '1662076800': 2, + '1662940800': 2, + '1663027200': 2, + '1663113600': 2, + '1663286400': 4, + '1663632000': 2, + '1664755200': 2, + '1664841600': 3, + '1665014400': 3, + '1665100800': 2, + '1665360000': 2, + '1665619200': 3, + '1666224000': 2, + '1667174400': 2, + '1667260800': 2, + '1668124800': 2, + '1668643200': 2, + '1668729600': 2, + '1669075200': 2, + '1669852800': 6, + '1670025600': 2, + '1670198400': 5, + '1670284800': 4, + '1670371200': 3, + '1670457600': 2, + '1670544000': 3, + '1670803200': 2, + '1670976000': 5, + '1671148800': 3, + '1671408000': 4, + '1671494400': 4, + '1671580800': 9, + '1671667200': 5, + '1672272000': 2, + '1672444800': 2, + '1672617600': 6, + '1672704000': 4, + '1672876800': 3, + '1672963200': 5, + '1673049600': 2, + '1673136000': 2, + '1673222400': 9, + '1673308800': 5, + '1673395200': 3, + '1673481600': 4, + '1673568000': 4, + '1673654400': 2, + '1673827200': 3, + '1673913600': 12, + '1674000000': 5, + '1674086400': 4, + '1674172800': 4, + '1674432000': 2, + '1674518400': 3, + '1674604800': 4, + '1674691200': 3, + '1675036800': 5, + '1675123200': 8, + '1675209600': 7, + '1675382400': 3, + '1675641600': 6, + '1675728000': 6, + '1675814400': 16, + '1675900800': 6, + '1675987200': 4, + '1676246400': 8, + '1676332800': 7, + '1676419200': 9, + '1676505600': 20, + '1676592000': 6, + '1676678400': 5, + '1676764800': 5, + '1676851200': 6, + '1676937600': 14, + '1677024000': 3, + '1677110400': 3, + '1677196800': 3, + '1677456000': 3, + '1677542400': 2, + '1677715200': 2, + '1677974400': 2, + '1678060800': 7, + '1678147200': 4, + '1678233600': 2, + '1678492800': 3, + '1678665600': 8, + '1678838400': 2, + '1678924800': 3, + '1679011200': 3, + '1679356800': 4, + '1679443200': 3, + '1679529600': 4, + '1679616000': 2, + '1679961600': 7, + '1680220800': 5, + '1680566400': 2, + '1680652800': 3, + '1681344000': 3, + '1681862400': 4, + '1682121600': 2, + '1682380800': 2, + '1682467200': 3, + '1682553600': 4, + '1683763200': 5, + '1683849600': 9, + '1684195200': 2, + '1684281600': 4, + '1684368000': 5, + '1684454400': 2, + '1684713600': 10, + '1684800000': 9, + '1684972800': 3, + '1685145600': 10, + '1685232000': 3, + '1685318400': 9, + '1685404800': 33, + '1685491200': 15, + '1685577600': 20, + '1685664000': 10, + '1685923200': 4, + '1686096000': 5, + '1686182400': 6, + '1686268800': 2, + '1686614400': 3, + '1687392000': 2, + '1687478400': 3, + '1687737600': 2, + '1687910400': 4, + '1688083200': 2, + '1688774400': 3, + '1689033600': 3, + '1689120000': 2, + '1690156800': 2, + '1693353600': 3, + '1693612800': 2, + '1694563200': 3, + '1695168000': 2, + '1695600000': 2, + '1695859200': 4, + '1696204800': 9, + '1696291200': 8, + '1696377600': 4, + '1696464000': 2, + '1696809600': 7, + '1697500800': 2, + '1698019200': 2, + '1698105600': 6, + '1698278400': 3, + '1698710400': 8, + '1698796800': 4, + '1698883200': 17, + '1698969600': 13, + '1699056000': 3, + '1699228800': 7, + '1699315200': 3, + '1699401600': 9, + '1699488000': 6, + '1699747200': 2, + '1699920000': 3, + '1700006400': 6, + '1700092800': 9, + '1700179200': 2, + '1700265600': 2, + '1700438400': 2, + '1700697600': 2, + '1701043200': 2, + '1701129600': 11, + '1701820800': 2, + '1701907200': 2, + '1702252800': 4, + '1702339200': 20, + '1702512000': 3, + '1702857600': 3, + '1702944000': 6, + '1703030400': 15, + '1703116800': 5, + '1703548800': 3, + '1703635200': 4, + '1703721600': 2, + '1703808000': 11, + '1704067200': 2, + '1704758400': 4, + '1704844800': 8, + '1704931200': 8, + '1705017600': 6, + '1705104000': 4, + '1705276800': 4, + '1705363200': 7, + '1705622400': 2, + '1705881600': 5, + '1705968000': 2, + '1706054400': 4, + '1706140800': 3, + '1706227200': 6, + '1706486400': 2, + '1706572800': 16, + '1706659200': 14, + '1706832000': 3, + '1707004800': 4, + '1707091200': 14, + '1707177600': 5, + '1707264000': 2, + '1707350400': 4, + '1707696000': 6, + '1707782400': 9, + '1707868800': 12, + '1707955200': 8, + '1708041600': 3, + '1708128000': 2, + '1708214400': 2, + '1708300800': 4, + '1708473600': 14, + '1708560000': 26, + '1708646400': 9, + '1708732800': 3, + '1708905600': 4, + '1708992000': 3, + '1709251200': 2, + '1709856000': 2, + '1710374400': 5, + '1711152000': 3, + '1711497600': 3, + '1711584000': 2, + '1711670400': 2, + '1712016000': 3, + '1712102400': 8, + '1712188800': 3, + '1712275200': 4, + '1712448000': 2, + '1712534400': 3, + '1712707200': 2, + '1713139200': 2, + '1713830400': 2, + '1713916800': 2, + '1714003200': 8, + '1714089600': 7, + '1714176000': 3, + '1714348800': 2, + '1715644800': 3, + '1715731200': 4, + '1715817600': 2, + '1715904000': 4, + '1716163200': 12, + '1716249600': 6, + '1716336000': 4, + '1716422400': 7, + '1716508800': 3, + '1716595200': 2, + '1716681600': 2, + '1716768000': 2, + '1717113600': 2, + '1717200000': 2, + '1717977600': 3, + '1718064000': 3, + '1718150400': 5, + '1718236800': 4, + '1718668800': 2, + '1718841600': 3, + '1718928000': 2, + '1719187200': 4, + '1719273600': 2, + '1719532800': 2, + '1719619200': 3, + '1719792000': 2, + '1719878400': 3, + '1719964800': 2, + '1720569600': 2, + '1721692800': 2, + '1721952000': 2, + '1722124800': 2, + '1722384000': 2, + '1723507200': 2, + '1723593600': 6, + '1724025600': 3, + '1724112000': 2, + '1724198400': 2, + '1724716800': 2, + '1724976000': 2, + '1725235200': 5, + '1725321600': 5, + '1725408000': 4, + '1726012800': 2, + '1726185600': 2, + '1726704000': 2, + '1727136000': 2, + '1727222400': 3, + '1727308800': 10, + '1727740800': 3, + '1727827200': 6, + '1727913600': 2, + '1728000000': 2, + '1728086400': 3, + '1728172800': 2, + '1728259200': 4, + '1728432000': 2, + '1728604800': 10, + '1728691200': 6, + '1728777600': 3, + '1728864000': 18, + '1728950400': 7, + '1729036800': 3, + '1729123200': 5, + '1729209600': 7, + '1729468800': 4, + '1729555200': 8, + '1729641600': 12, + '1729728000': 15, + '1729814400': 5, + '1729900800': 3, + '1732060800': 3, + }, + 'CSETv1.Autonomy Level': { + Autonomy1: 847, + Autonomy3: 370, + Autonomy2: 261, + unclear: 104, + }, + 'CSETv1.Involving Minor': { + no: 1469, + yes: 106, + maybe: 7, + }, + 'CSETv1.Location Region': { + 'North America': 781, + Global: 402, + Asia: 177, + Europe: 152, + Oceania: 64, + Africa: 5, + 'South America': 1, + }, + 'CSETv1.Intentional Harm': { + 'No. Not intentionally designed to perform harm': 1462, + 'Yes. Intentionally designed to perform harm and did create intended harm': 70, + unclear: 44, + 'Yes. Intentionally designed to perform harm but created an unintended harm (a different harm may have occurred)': 2, + }, + 'CSETv1.Physical Objects': { + no: 946, + yes: 615, + maybe: 21, + }, + 'CSETv1.Rights Violation': { + no: 1332, + yes: 168, + maybe: 82, + }, + 'GMF.Known AI Technology': { + 'Face Detection': 69, + 'Recurrent Neural Network': 55, + 'Distributional Learning': 53, + 'Convolutional Neural Network': 50, + 'Language Modeling': 43, + 'Automatic Speech Recognition': 37, + 'Acoustic Fingerprint': 35, + 'Neural Network': 35, + 'Generative Adversarial Network': 33, + 'Keyword Filtering': 24, + 'Geolocation Data': 15, + 'Collaborative Filtering': 14, + 'Content-based Filtering': 14, + Transformer: 14, + Autoencoder: 13, + 'Acoustic Triangulation': 11, + 'Character NGrams': 9, + Classification: 6, + 'Gesture Recognition': 6, + 'Diverse Data': 4, + 'Multimodal Learning': 4, + Regression: 4, + 'Image Segmentation': 2, + 'Optical Character Recognition': 1, + 'Visual Object Detection': 1, + }, + 'CSETv1.Detrimental Content': { + no: 1281, + yes: 262, + maybe: 39, + }, + 'CSETv1.AI tools and methods': { + 'natural language processing': 235, + 'computer vision': 217, + 'Natural Language Processesing': 77, + 'facial recognition': 55, + 'voice recognition': 55, + 'neural networks': 54, + 'audio transcription': 51, + 'machine learning': 51, + 'image mapping': 45, + 'point mapping': 45, + unclear: 42, + 'object detection': 38, + 'human language technology': 32, + 'search engine optimization': 31, + 'Sensor Data Processing': 29, + prediction: 28, + 'Dijkstra Algorithm': 24, + 'face detection': 24, + 'logistic regression': 24, + 'shortest-path algorithm': 24, + 'non-linear SVM': 22, + 'signal detection theory': 22, + 'facial reconstruction': 21, + 'image reconstruction': 21, + 'large language models': 17, + 'image match': 16, + 'text to speech': 16, + 'deep learning': 15, + 'mesh autoencoders': 13, + classification: 12, + 'vector embedding': 11, + 'machine vision': 10, + 'large language model': 8, + 'regular expressions': 6, + robotics: 5, + 'speech recognition': 5, + estimation: 4, + 'neural network': 3, + transformer: 3, + 'Text to Speach': 2, + 'emotion detection': 2, + 'large-language models': 2, + 'logistic regression model': 2, + ranking: 2, + 'C4.5 algorithm': 1, + 'GPT-3': 1, + 'Markov decision process': 1, + 'SSD network algorithm': 1, + 'VCG network': 1, + 'YOLO network algorithm': 1, + classiification: 1, + 'convolutional neural networks': 1, + 'decision tree': 1, + 'linear algebra': 1, + 'natural language': 1, + 'natural language response': 1, + 'pixel embeddings': 1, + recommendation: 1, + 'reinforcement learning': 1, + 'response model': 1, + 'text scraping': 1, + 'unsupervised learning': 1, + 'website code': 1, + }, + 'CSETv1.Operating Conditions': { + 'operationally representative': 63, + Testing: 29, + 'Driverless car': 28, + night: 25, + 'natural disaster - wildfires': 22, + 'Twin faces': 21, + 'unclear enunciation': 16, + 'inclement weather and snow': 6, + 'COVID-19 pandemic': 4, + 'inclement weather - snow': 2, + 'seasonal road closures': 2, + 'Non-operational': 1, + }, + 'CSETv1.Physical System Type': { + 'Manufacturing Robot': 67, + none: 59, + 'rocket/egg shaped, 300 pound, 5 ft. tall security robot': 57, + 'Apple iPhone X': 45, + 'Amazon Alexa Echo Dot speaker': 35, + 'Volvo XC90 SUV': 35, + Camera: 29, + 'Metro train': 29, + 'Vehicles (Lexus, Audi, Chrysler Pacifica)': 28, + 'Tesla vehicle (Model 3, Model S, Model X)': 25, + 'Tesla Model S': 24, + 'oval, eight-seater, autonomous, electric shuttle bus': 24, + 'Boeing 737 Max airplane': 19, + 'Amazon Echo Dot Smart Speaker': 16, + 'medical robot': 12, + 'Vehicles (Audi Q5 Crossover, Lexus RX400h Crossover)': 11, + microphones: 11, + 'Electric vehicle': 10, + 'Tesla Model 3': 6, + 'smoke detector': 6, + 'A cone-shaped, 400-pound security robot on wheels': 5, + 'Nissan vehicles (Rogue, Rogue Hybrid, Rogue Sport, Sentra)': 5, + 'Amazon Echo': 4, + 'Tesla sedan': 4, + 'Vehicles (Volvo, Lexus, Ford, Google, etc.)': 4, + drone: 4, + 'fulfillment robot': 4, + '2016 Tesla Model S': 3, + 'Handheld barcode scanner; flat, round mobile robots;': 3, + '4ft tall vaguely humanoid droid on wheels': 2, + 'Amazon Echo Dot': 2, + 'Humanoid robot': 2, + 'body scanner': 2, + car: 2, + 'store security camera': 2, + 'Palm VII': 1, + 'Starship delivery robot': 1, + 'in-van cameras and sensors': 1, + 'robotic dog': 1, + 'vaguely humanoid robot': 1, + }, + 'CSETv1.Sector of Deployment': { + 'information and communication': 562, + 'transportation and storage': 273, + 'Arts, entertainment and recreation': 190, + 'wholesale and retail trade': 154, + 'administrative and support service activities': 125, + 'public administration': 117, + 'law enforcement': 114, + 'human health and social work activities': 92, + 'professional, scientific and technical activities': 71, + 'financial and insurance activities': 69, + manufacturing: 67, + Education: 41, + defense: 31, + 'accommodation and food service activities': 18, + other: 10, + 'real estate activities': 5, + 'other service activities': 2, + unclear: 2, + }, + 'GMF.Potential AI Technology': { + Regression: 54, + 'Image Segmentation': 45, + Classification: 44, + 'Convolutional Neural Network': 44, + 'Ensemble Aggregation': 39, + 'Distributional Learning': 35, + '3D reconstruction': 29, + 'Other domain-specific approaches': 29, + 'Face Detection': 27, + 'Satellite Imaging': 27, + 'Image Classification': 26, + 'Intermediate modeling': 26, + 'Multimodal Learning': 26, + 'Diverse Data': 13, + 'Visual Object Detection': 11, + Clustering: 6, + 'Geolocation Data': 4, + 'Optical Character Recognition': 4, + Transformer: 3, + 'Acoustic Fingerprint': 2, + Spectrogram: 2, + 'Siamese Network': 1, + }, + 'CSETv1.Date of Incident Year': { + '1979': 4, + '1983': 27, + '2000': 13, + '2001': 6, + '2008': 12, + '2009': 28, + '2010': 30, + '2011': 43, + '2012': 41, + '2013': 53, + '2014': 95, + '2015': 78, + '2016': 268, + '2017': 382, + '2018': 181, + '2019': 43, + '2020': 83, + '2021': 82, + '2022': 36, + '2023': 77, + }, + 'CSETv1.Entertainment Industry': { + no: 1401, + yes: 140, + maybe: 41, + }, + 'CSETv1.Infrastructure Sectors': { + transportation: 139, + 'healthcare and public health': 75, + 'financial services': 36, + 'defense-industrial base': 31, + communications: 29, + 'information technology': 25, + 'emergency services': 23, + 'food and agriculture': 8, + Other: 3, + 'commercial facilities': 2, + 'government facilities': 2, + unclear: 2, + }, + 'CSETv1.Harm Distribution Basis': { + none: 1059, + race: 346, + sex: 198, + 'nation of origin, citizenship, immigrant status': 100, + religion: 88, + 'sexual orientation or gender identity': 73, + 'financial means': 44, + disability: 41, + geography: 35, + ideology: 33, + age: 30, + 'familial status (e.g., having or not having children) or pregnancy': 4, + unclear: 4, + other: 3, + }, + 'CSETv1.Multiple AI Interaction': { + no: 1525, + yes: 48, + maybe: 9, + }, + 'GMF.Known AI Technical Failure': { + 'Unsafe Exposure or Access': 59, + 'Distributional Bias': 55, + Misuse: 37, + 'Adversarial Data': 29, + 'Generalization Failure': 28, + 'Limited Dataset': 28, + 'Black Swan Event': 27, + 'Data or Labelling Noise': 27, + 'Dataset Imbalance': 27, + Underfitting: 24, + 'Lack of Adversarial Robustness': 23, + 'Tuning Issues': 17, + 'Inappropriate Training Content': 16, + 'Inadequate Anonymization': 13, + 'Unauthorized Data': 13, + Underspecification: 10, + 'Context Misidentification': 9, + 'Gaming Vulnerability': 5, + 'Lack of Capability Control': 5, + 'Concept Drift': 4, + 'Misinformation Generation Hazard': 3, + 'Incomplete Data Attribute Capture': 2, + 'Lack of Explainability': 2, + 'Lack of Transparency': 2, + 'Misconfigured Threshold': 1, + }, + 'CSETv1.Clear link to Technology': { + yes: 1425, + no: 105, + maybe: 52, + }, + 'CSETv1.Protected Characteristic': { + no: 1070, + yes: 505, + maybe: 7, + }, + 'CSETv1.Public Sector Deployment': { + no: 1324, + yes: 249, + maybe: 9, + }, + 'CSETv1.Impact on Critical Services': { + no: 1474, + yes: 106, + maybe: 2, + }, + 'GMF.Potential AI Technical Failure': { + 'Generalization Failure': 135, + 'Context Misidentification': 94, + 'Dataset Imbalance': 56, + Underspecification: 52, + 'Unauthorized Data': 45, + Underfitting: 41, + 'Hardware Failure': 39, + 'Lack of Capability Control': 39, + 'Data or Labelling Noise': 38, + 'Inadequate Anonymization': 35, + Overfitting: 34, + 'Misinformation Generation Hazard': 33, + 'Gaming Vulnerability': 30, + 'Distributional Bias': 29, + Misuse: 29, + 'Misconfigured Aggregation': 26, + 'Concept Drift': 25, + 'Misaligned Objective': 23, + 'Misconfigured Threshold': 18, + 'Incomplete Data Attribute Capture': 17, + 'Limited Dataset': 17, + 'Covariate Shift': 16, + 'Inadequate Data Sampling': 13, + 'Tuning Issues': 10, + 'Limited User Access': 6, + 'Data Memorization': 5, + 'Lack of Explainability': 5, + 'Adversarial Data': 4, + 'Backup Failure': 4, + 'Lack of Transparency': 4, + 'Software Bug': 4, + 'Robustness Failure': 3, + 'Black Box': 2, + 'Lack of Adversarial Robustness': 2, + 'Problematic Input': 2, + }, + 'CSETv1.Location Country (two letters)': { + US: 769, + RU: 59, + CN: 45, + IN: 41, + AU: 40, + IE: 35, + DE: 32, + PS: 26, + VN: 26, + NZ: 24, + ID: 19, + KR: 17, + GB: 13, + NL: 6, + 'United States': 6, + CA: 5, + FR: 5, + GR: 4, + LY: 4, + SE: 4, + IT: 2, + JP: 2, + AR: 1, + CH: 1, + IL: 1, + RS: 1, + }, + 'CSETv1.Report, Test, or Study of data': { + no: 1511, + yes: 63, + maybe: 8, + }, + 'CSETv1.Special Interest Intangible Harm': { + no: 862, + yes: 703, + maybe: 17, + }, + 'CSETv1.User Test in Controlled Conditions': { + no: 1555, + yes: 25, + maybe: 2, + }, + 'CSETv1.User Test in Operational Conditions': { + no: 1367, + yes: 211, + maybe: 4, + }, + 'CSETv1.Producer Test in Controlled Conditions': { + no: 1509, + yes: 69, + maybe: 4, + }, + 'CSETv1.Producer Test in Operational Conditions': { + no: 1453, + yes: 129, + }, + 'CSETv1.Annotator’s AI special interest intangible harm assessment': { + no: 949, + yes: 609, + maybe: 24, + }, + }, + facets_stats: { + incident_id: { + min: 1, + max: 850, + avg: 336, + sum: 1304307, + }, + 'CSETv1.Injuries': { + min: 0, + max: 55000, + avg: 114, + sum: 184244, + }, + 'CSETv1.Lives Lost': { + min: 0, + max: 189, + avg: 3, + sum: 5922, + }, + epoch_incident_date: { + min: 433382400, + max: 1732060800, + avg: 1570344663, + sum: 6078804192000, + }, + epoch_date_published: { + min: 828489600, + max: 1732147200, + avg: 1606384020, + sum: 6218312544000, + }, + 'CSETv1.Date of Incident Year': { + min: 1979, + max: 2023, + avg: 2015, + sum: 3189067, + }, + }, + exhaustiveFacetsCount: true, + exhaustiveFacetValues: false, + exhaustiveNbHits: true, + exhaustive: { + facetsCount: true, + facetValues: false, + nbHits: true, + }, + query: '', + params: + 'distinct=false&facetFilters=%5B%5B%22is_incident_report%3Atrue%22%5D%5D&facets=%5B%22CSETv1.AI%20Harm%20Level%22%2C%22CSETv1.AI%20System%22%2C%22CSETv1.AI%20System%20Description%22%2C%22CSETv1.AI%20Tangible%20Harm%20Level%20Notes%22%2C%22CSETv1.AI%20Task%22%2C%22CSETv1.AI%20tools%20and%20methods%22%2C%22CSETv1.Annotation%20Status%22%2C%22CSETv1.Annotator%22%2C%22CSETv1.Annotator%E2%80%99s%20AI%20special%20interest%20intangible%20harm%20assessment%22%2C%22CSETv1.Autonomy%20Level%22%2C%22CSETv1.Clear%20link%20to%20Technology%22%2C%22CSETv1.Clear%20link%20to%20technology%22%2C%22CSETv1.Data%20Inputs%22%2C%22CSETv1.Date%20of%20Incident%20Day%22%2C%22CSETv1.Date%20of%20Incident%20Month%22%2C%22CSETv1.Date%20of%20Incident%20Year%22%2C%22CSETv1.Deployed%22%2C%22CSETv1.Detrimental%20Content%22%2C%22CSETv1.Embedded%22%2C%22CSETv1.Entertainment%20Industry%22%2C%22CSETv1.Entities%22%2C%22CSETv1.Estimated%20Date%22%2C%22CSETv1.Estimated%20Harm%20Quantities%22%2C%22CSETv1.Harm%20Distribution%20Basis%22%2C%22CSETv1.Harm%20Domain%22%2C%22CSETv1.Harmed%20Class%20of%20Entities%22%2C%22CSETv1.Impact%20on%20Critical%20Services%22%2C%22CSETv1.Incident%20Number%22%2C%22CSETv1.Infrastructure%20Sectors%22%2C%22CSETv1.Injuries%22%2C%22CSETv1.Intentional%20Harm%22%2C%22CSETv1.Involving%20Minor%22%2C%22CSETv1.Lives%20Lost%22%2C%22CSETv1.Location%20City%22%2C%22CSETv1.Location%20Country%20(two%20letters)%22%2C%22CSETv1.Location%20Region%22%2C%22CSETv1.Location%20State%2FProvince%20(two%20letters)%22%2C%22CSETv1.Multiple%20AI%20Interaction%22%2C%22CSETv1.Notes%20(%20Tangible%20Harm%20Quantities%20Information)%22%2C%22CSETv1.Notes%20(AI%20Functionality%20and%20Techniques)%22%2C%22CSETv1.Notes%20(AI%20special%20interest%20intangible%20harm)%22%2C%22CSETv1.Notes%20(Environmental%20and%20Temporal%20Characteristics)%22%2C%22CSETv1.Notes%20(Information%20about%20AI%20System)%22%2C%22CSETv1.Notes%20(special%20interest%20intangible%20harm)%22%2C%22CSETv1.Operating%20Conditions%22%2C%22CSETv1.Peer%20Reviewer%22%2C%22CSETv1.Physical%20Objects%22%2C%22CSETv1.Physical%20System%20Type%22%2C%22CSETv1.Producer%20Test%20in%20Controlled%20Conditions%22%2C%22CSETv1.Producer%20Test%20in%20Operational%20Conditions%22%2C%22CSETv1.Protected%20Characteristic%22%2C%22CSETv1.Public%20Sector%20Deployment%22%2C%22CSETv1.Quality%20Control%22%2C%22CSETv1.Report%2C%20Test%2C%20or%20Study%20of%20data%22%2C%22CSETv1.Rights%20Violation%22%2C%22CSETv1.Sector%20of%20Deployment%22%2C%22CSETv1.Special%20Interest%20Intangible%20Harm%22%2C%22CSETv1.Tangible%20Harm%22%2C%22CSETv1.There%20is%20a%20potentially%20identifiable%20specific%20entity%20that%20experienced%20the%20harm%22%2C%22CSETv1.User%20Test%20in%20Controlled%20Conditions%22%2C%22CSETv1.User%20Test%20in%20Operational%20Conditions%22%2C%22GMF.Known%20AI%20Goal%22%2C%22GMF.Known%20AI%20Goal%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Goal%20Snippets%22%2C%22GMF.Known%20AI%20Technical%20Failure%22%2C%22GMF.Known%20AI%20Technical%20Failure%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Technical%20Failure%20Snippets%22%2C%22GMF.Known%20AI%20Technology%22%2C%22GMF.Known%20AI%20Technology%20Classification%20Discussion%22%2C%22GMF.Known%20AI%20Technology%20Snippets%22%2C%22GMF.Potential%20AI%20Goal%22%2C%22GMF.Potential%20AI%20Goal%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Goal%20Snippets%22%2C%22GMF.Potential%20AI%20Technical%20Failure%22%2C%22GMF.Potential%20AI%20Technical%20Failure%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Technical%20Failure%20Snippets%22%2C%22GMF.Potential%20AI%20Technology%22%2C%22GMF.Potential%20AI%20Technology%20Classification%20Discussion%22%2C%22GMF.Potential%20AI%20Technology%20Snippets%22%2C%22authors%22%2C%22classifications%22%2C%22epoch_date_published%22%2C%22epoch_incident_date%22%2C%22flag%22%2C%22incident_id%22%2C%22is_incident_report%22%2C%22language%22%2C%22namespaces%22%2C%22source_domain%22%2C%22submitters%22%2C%22tags%22%5D&highlightPostTag=__%2Fais-highlight__&highlightPreTag=__ais-highlight__&hitsPerPage=28&maxValuesPerFacet=999&page=0&query=&tagFilters=', + index: 'instant_search-en-featured', + renderingContent: {}, + processingTimeMS: 3, + processingTimingsMS: { + _request: { + roundTrip: 160, + }, + afterFetch: { + format: { + highlighting: 3, + snippeting: 6, + total: 12, + }, + total: 2, + }, + fetch: { + total: 1, + }, + total: 3, + }, + serverTimeMS: 16, + }, + ], +}; + diff --git a/site/gatsby-site/playwright/utils.ts b/site/gatsby-site/playwright/utils.ts index 66ac5c8893..0391f610e5 100644 --- a/site/gatsby-site/playwright/utils.ts +++ b/site/gatsby-site/playwright/utils.ts @@ -11,6 +11,7 @@ import { ObjectId } from 'bson'; import users from './seeds/customData/users'; import authUsers from './seeds/auth/users'; import { algoliaMock } from './fixtures/algoliaMock'; +import siteConfig from '../config'; declare module '@playwright/test' { interface Request { @@ -25,6 +26,8 @@ type TestFixtures = { runOnlyOnEmptyEnvironment: () => Promise, login: (options?: { email?: string, customData?: Record }) => Promise, retryDelay?: [({ }: {}, use: () => Promise, testInfo: { retry: number }) => Promise, { auto: true }], + runOnlyInProduction: () => Promise, + runAnywhereExceptProduction: () => Promise, }; export function randomString(size: number) { @@ -159,6 +162,22 @@ export const test = base.extend({ await use(); }, { auto: true }], + + runOnlyInProduction: async ({ }, use, testInfo) => { + if (config.SITE_URL !== siteConfig.gatsby.siteUrl) { + testInfo.skip(); + } + + await use(null); + }, + + runAnywhereExceptProduction: async ({ }, use, testInfo) => { + if (config.SITE_URL === siteConfig.gatsby.siteUrl) { + testInfo.skip(); + } + + await use(null); + } }); // SEE: https://playwright.dev/docs/api/class-page#page-wait-for-request diff --git a/site/gatsby-site/server/emails/templates/EntityIncidentUpdated.ts b/site/gatsby-site/server/emails/templates/EntityIncidentUpdated.ts index 4653d01702..750a5f344f 100644 --- a/site/gatsby-site/server/emails/templates/EntityIncidentUpdated.ts +++ b/site/gatsby-site/server/emails/templates/EntityIncidentUpdated.ts @@ -165,7 +165,7 @@ export default `Incident Description: {{incidentDescription}}
Incident Date: {{incidentDate}}

-
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. The implicated systems are {{implicatedSystems}}.
+
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. Implicated AI system(s): {{implicatedSystems}}.

Sincerely,
Responsible AI Collaborative
diff --git a/site/gatsby-site/server/emails/templates/NewEntityIncident.ts b/site/gatsby-site/server/emails/templates/NewEntityIncident.ts index c033435c78..df8f73acd1 100644 --- a/site/gatsby-site/server/emails/templates/NewEntityIncident.ts +++ b/site/gatsby-site/server/emails/templates/NewEntityIncident.ts @@ -166,7 +166,7 @@ export default `Incident Description: {{incidentDescription}}
Incident Date: {{incidentDate}}

-
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. The implicated systems are {{implicatedSystems}}.
+
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. Implicated AI system(s): {{implicatedSystems}}.

Sincerely,
Responsible AI Collaborative
diff --git a/site/gatsby-site/server/emails/templates/NewIncident.ts b/site/gatsby-site/server/emails/templates/NewIncident.ts index c49d028092..0f54e24a41 100644 --- a/site/gatsby-site/server/emails/templates/NewIncident.ts +++ b/site/gatsby-site/server/emails/templates/NewIncident.ts @@ -165,7 +165,7 @@ export default `Incident Description: {{incidentDescription}}
Incident Date: {{incidentDate}}

-
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. The implicated systems are {{implicatedSystems}}.
+
Alleged: {{developers}} developed an AI system deployed by {{deployers}}, which harmed {{entitiesHarmed}}. Implicated AI system(s): {{implicatedSystems}}.

Sincerely,
Responsible AI Collaborative
diff --git a/site/gatsby-site/src/components/discover/hitTypes/Compact.js b/site/gatsby-site/src/components/discover/hitTypes/Compact.js index 6085d41c34..859855e83b 100644 --- a/site/gatsby-site/src/components/discover/hitTypes/Compact.js +++ b/site/gatsby-site/src/components/discover/hitTypes/Compact.js @@ -68,7 +68,12 @@ export default function Compact({ item, toggleFilterByIncidentId, viewType }) { /> - + {item.is_translated && ( + + )} diff --git a/site/gatsby-site/src/components/discover/hitTypes/Details.js b/site/gatsby-site/src/components/discover/hitTypes/Details.js index 5e3dc76cf4..c4c4c7d937 100644 --- a/site/gatsby-site/src/components/discover/hitTypes/Details.js +++ b/site/gatsby-site/src/components/discover/hitTypes/Details.js @@ -53,7 +53,9 @@ export default function Details({ item, toggleFilterByIncidentId, viewType }) { - + {item.is_translated && ( + + )} - -
- +
+
+
- + {item.is_translated && ( +
+ +
+ )} +
); return ; diff --git a/site/gatsby-site/src/components/entities/AllegedEntities.js b/site/gatsby-site/src/components/entities/AllegedEntities.js index 0d40495ccf..b3d193c742 100644 --- a/site/gatsby-site/src/components/entities/AllegedEntities.js +++ b/site/gatsby-site/src/components/entities/AllegedEntities.js @@ -51,11 +51,8 @@ function ImplicatedSystemsList({ entities }) {
{entities.length > 0 && ( <> - - The AI implicated system{entities.length > 1 ? 's' : ''}{' '} - {entities.length > 1 ? 'are' : 'is'} - {' '} - . + Implicated AI system{entities.length > 1 ? 's' : ''}:{' '} + {' '} )}
diff --git a/site/gatsby-site/src/utils/AlgoliaUpdater.js b/site/gatsby-site/src/utils/AlgoliaUpdater.js index d9d54386ca..d044ba082c 100644 --- a/site/gatsby-site/src/utils/AlgoliaUpdater.js +++ b/site/gatsby-site/src/utils/AlgoliaUpdater.js @@ -135,6 +135,7 @@ const reportToEntry = ({ incident = null, report, classifications = [{ list: [], featured: featuredValue, flag: report.flag, is_incident_report: report.is_incident_report, + is_translated: report.is_translated, namespaces: classifications.map((c) => { return Object.keys(c.tree)[0]; }), @@ -300,8 +301,13 @@ class AlgoliaUpdater { .toArray(); const fullReports = reports.map((r) => { - let report = { ...r }; + // by default, use the report as is + let report = { + ...r, + is_translated: false, + }; + // If the report has a translation, use it if (translations.some((t) => t.report_number === r.report_number)) { const { title, plain_text } = translations.find((t) => t.report_number === r.report_number) || {}; @@ -310,6 +316,7 @@ class AlgoliaUpdater { ...r, title, plain_text, + is_translated: true, }; } return report;