When scheduling multiple events, the admin UI sends a request to check for conflicts. It looks like there is a timezone issue within the generated RRULE, as local time is used in BYHOUR while UTC is used for the start and end timestamps:

POST https://mec.admin.opencast.tum.de/admin-ng/event/new/conflicts
{
    "metadata": "{\"start\":\"2024-10-13T06:10:00.000Z\",\"device\":\"MRI+-+HS-B\",\"duration\":\"6300000\",\"end\":\"2025-01-01T08:55:00.000Z\",\"rrule\":\"FREQ=WEEKLY;BYDAY=TU;BYHOUR=8;BYMINUTE=10\"}"
}

The old admin UI on the other hand used UTC consistently:

{
    "metadata": "{\"start\":\"2024-10-13T06:10:00Z\",\"device\":\"MRI - HS-B\",\"duration\":\"6300000\",\"end\":\"2025-01-01T08:55:00Z\",\"rrule\":\"FREQ=WEEKLY;BYDAY=TU;BYHOUR=6;BYMINUTE=10\"}"
}

As a result, wrong (i.e. non-existing) scheduling conflicts are reported in the UI, preventing event creation. (On the other hand, actual conflicts may remain undetected, but I have not tested this.)

EDIT:

Maybe getUTCHours and getUTCMinutes should be used here?