diff --git a/NEWS.md b/NEWS.md
index 178d50e72..145cbecb6 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -1,6 +1,16 @@
 # targets 1.3.2.9001 (development)
 
-* Add function `tar_seed_set()` which sets a seed and the default RNG algorithms.
+## Invalidating changes
+
+Because of the changes below, upgrading to this version of `targets` will unavoidably invalidate previously built targets in existing pipelines. Your pipeline code should still work, but any targets you ran before will most likely need to rerun after the upgrade.
+
+* Use SHA512 during the creation of target-specific pseudo-random number generator seeds (#1139). This change decreases the risk of overlapping/correlated random number generator streams. See the "RNG overlap" section of the `tar_seed_create()` help file for details and justification.
+
+## Other changes
+
+* Add a new exported function `tar_seed_create()` which creates target-specific pseudo-random number generator seeds.
+* Add an "RNG overlap" section in the `tar_seed_create()` help file to justify and defend how `targets` and `tarchetypes` approach pseudo-random numbers.
+* Add function `tar_seed_set()` which sets a seed and sets all the RNG algorithms to their defaults in the R installation of the user. Each target now uses `tar_seed_set()` function to set its seed before running its R command (#1139).
 * Deprecate `tar_seed()` in favor of the new `tar_seed_get()` function.
 
 # targets 1.3.2
diff --git a/R/class_branch.R b/R/class_branch.R
index f0949c1c7..fbe747ed1 100644
--- a/R/class_branch.R
+++ b/R/class_branch.R
@@ -10,7 +10,7 @@ branch_init <- function(
   command <- command_clone(command)
   deps <- union(command$deps, deps)
   command$deps <- setdiff(deps, settings$dimensions)
-  command$seed <- produce_seed(child)
+  command$seed <- tar_seed_create(child)
   pedigree <- pedigree_new(settings$name, child, index)
   settings <- settings_clone(settings)
   settings$name <- child
diff --git a/R/class_target.R b/R/class_target.R
index 115cd44ce..a594e5182 100644
--- a/R/class_target.R
+++ b/R/class_target.R
@@ -19,7 +19,7 @@ target_init <- function(
   retrieval = "main",
   cue = NULL
 ) {
-  seed <- produce_seed(name)
+  seed <- tar_seed_create(name)
   command <- command_init(expr, packages, library, seed, deps, string)
   cue <- cue %|||% cue_init()
   if (any(grepl("^aws_", format))) {
diff --git a/R/tar_seed_create.R b/R/tar_seed_create.R
index d1a3284d4..da23b152b 100644
--- a/R/tar_seed_create.R
+++ b/R/tar_seed_create.R
@@ -4,8 +4,8 @@
 #' @description Create a seed for a target.
 #' @section Seeds:
 #'   A target's random number generator seed
-#'   is a deterministic function of its name and the global pipeline seed.
-#'   Consequently,
+#'   is a deterministic function of its name and the global pipeline seed
+#'   from [tar_option_get("seed")]. Consequently,
 #'
 #'     1. Each target runs with a reproducible seed so that
 #'        different runs of the same pipeline in the same computing
@@ -24,14 +24,14 @@
 #'   correlated results. (For a discussion of the motivating problem,
 #'   see the Section 6: "Random-number generation" in the `parallel`
 #'   package vignette: `vignette(topic = "parallel", package = "parallel")`.)
-#'   However, this risk is extremely small in practice.
-#'   
-#'   `targets` and `tarchetypes` take the approach discussed in
+#'   However, this risk is extremely small in practice, as shown by
 #'   L'Ecuyer et al. (2027) <https://doi.org/10.1016/j.matcom.2016.05.005>
-#'   "A single RNG with a 'random' seed for each stream" (Section 4:
+#'   under "A single RNG with a 'random' seed for each stream" (Section 4:
 #'   under "How to produce parallel streams and substreams").
-#'   Here, [tar_seed_create()] plays the role
-#'   of the upstream pseudo-random number generator (RNG) that produces
+#'
+#'   `targets` and `tarchetypes` take the approach discussed in the
+#'   aforementioned section of the paper, where [tar_seed_create()] plays the
+#'   role of the upstream pseudo-random number generator (RNG) that produces
 #'   seeds for the subsequent parallel streams. Specifically,
 #'   [tar_seed_create()] acts as a counter-based RNG,
 #'   where the output function is the SHA512 hash algorithm.
diff --git a/man/tar_seed_create.Rd b/man/tar_seed_create.Rd
index 807992fae..02762a8dd 100644
--- a/man/tar_seed_create.Rd
+++ b/man/tar_seed_create.Rd
@@ -12,8 +12,8 @@ Create a seed for a target.
 \section{Seeds}{
 
 A target's random number generator seed
-is a deterministic function of its name and the global pipeline seed.
-Consequently,
+is a deterministic function of its name and the global pipeline seed
+from \link{tar_option_get("seed")}. Consequently,
 
 \if{html}{\out{<div class="sourceCode">}}\preformatted{1. Each target runs with a reproducible seed so that
    different runs of the same pipeline in the same computing
@@ -36,14 +36,14 @@ streams of different targets will overlap and produce statistically
 correlated results. (For a discussion of the motivating problem,
 see the Section 6: "Random-number generation" in the \code{parallel}
 package vignette: \code{vignette(topic = "parallel", package = "parallel")}.)
-However, this risk is extremely small in practice.
-
-\code{targets} and \code{tarchetypes} take the approach discussed in
+However, this risk is extremely small in practice, as shown by
 L'Ecuyer et al. (2027) \url{https://doi.org/10.1016/j.matcom.2016.05.005}
-"A single RNG with a 'random' seed for each stream" (Section 4:
+under "A single RNG with a 'random' seed for each stream" (Section 4:
 under "How to produce parallel streams and substreams").
-Here, \code{\link[=tar_seed_create]{tar_seed_create()}} plays the role
-of the upstream pseudo-random number generator (RNG) that produces
+
+\code{targets} and \code{tarchetypes} take the approach discussed in the
+aforementioned section of the paper, where \code{\link[=tar_seed_create]{tar_seed_create()}} plays the
+role of the upstream pseudo-random number generator (RNG) that produces
 seeds for the subsequent parallel streams. Specifically,
 \code{\link[=tar_seed_create]{tar_seed_create()}} acts as a counter-based RNG,
 where the output function is the SHA512 hash algorithm.
diff --git a/man/tar_seed_get.Rd b/man/tar_seed_get.Rd
index 82e0bd36c..77e33c553 100644
--- a/man/tar_seed_get.Rd
+++ b/man/tar_seed_get.Rd
@@ -25,8 +25,8 @@ of the target currently running.
 \section{Seeds}{
 
 A target's random number generator seed
-is a deterministic function of its name and the global pipeline seed.
-Consequently,
+is a deterministic function of its name and the global pipeline seed
+from \link{tar_option_get("seed")}. Consequently,
 
 \if{html}{\out{<div class="sourceCode">}}\preformatted{1. Each target runs with a reproducible seed so that
    different runs of the same pipeline in the same computing
@@ -49,14 +49,14 @@ streams of different targets will overlap and produce statistically
 correlated results. (For a discussion of the motivating problem,
 see the Section 6: "Random-number generation" in the \code{parallel}
 package vignette: \code{vignette(topic = "parallel", package = "parallel")}.)
-However, this risk is extremely small in practice.
-
-\code{targets} and \code{tarchetypes} take the approach discussed in
+However, this risk is extremely small in practice, as shown by
 L'Ecuyer et al. (2027) \url{https://doi.org/10.1016/j.matcom.2016.05.005}
-"A single RNG with a 'random' seed for each stream" (Section 4:
+under "A single RNG with a 'random' seed for each stream" (Section 4:
 under "How to produce parallel streams and substreams").
-Here, \code{\link[=tar_seed_create]{tar_seed_create()}} plays the role
-of the upstream pseudo-random number generator (RNG) that produces
+
+\code{targets} and \code{tarchetypes} take the approach discussed in the
+aforementioned section of the paper, where \code{\link[=tar_seed_create]{tar_seed_create()}} plays the
+role of the upstream pseudo-random number generator (RNG) that produces
 seeds for the subsequent parallel streams. Specifically,
 \code{\link[=tar_seed_create]{tar_seed_create()}} acts as a counter-based RNG,
 where the output function is the SHA512 hash algorithm.
diff --git a/man/tar_seed_set.Rd b/man/tar_seed_set.Rd
index 45da02632..08fd6d3a6 100644
--- a/man/tar_seed_set.Rd
+++ b/man/tar_seed_set.Rd
@@ -26,8 +26,8 @@ these seeds in R.
 \section{Seeds}{
 
 A target's random number generator seed
-is a deterministic function of its name and the global pipeline seed.
-Consequently,
+is a deterministic function of its name and the global pipeline seed
+from \link{tar_option_get("seed")}. Consequently,
 
 \if{html}{\out{<div class="sourceCode">}}\preformatted{1. Each target runs with a reproducible seed so that
    different runs of the same pipeline in the same computing
@@ -50,14 +50,14 @@ streams of different targets will overlap and produce statistically
 correlated results. (For a discussion of the motivating problem,
 see the Section 6: "Random-number generation" in the \code{parallel}
 package vignette: \code{vignette(topic = "parallel", package = "parallel")}.)
-However, this risk is extremely small in practice.
-
-\code{targets} and \code{tarchetypes} take the approach discussed in
+However, this risk is extremely small in practice, as shown by
 L'Ecuyer et al. (2027) \url{https://doi.org/10.1016/j.matcom.2016.05.005}
-"A single RNG with a 'random' seed for each stream" (Section 4:
+under "A single RNG with a 'random' seed for each stream" (Section 4:
 under "How to produce parallel streams and substreams").
-Here, \code{\link[=tar_seed_create]{tar_seed_create()}} plays the role
-of the upstream pseudo-random number generator (RNG) that produces
+
+\code{targets} and \code{tarchetypes} take the approach discussed in the
+aforementioned section of the paper, where \code{\link[=tar_seed_create]{tar_seed_create()}} plays the
+role of the upstream pseudo-random number generator (RNG) that produces
 seeds for the subsequent parallel streams. Specifically,
 \code{\link[=tar_seed_create]{tar_seed_create()}} acts as a counter-based RNG,
 where the output function is the SHA512 hash algorithm.