AWS Security ChangesHomeSearch

AWS clean-rooms: Clarify always-allowed literal value exception in custom analysis rule

Service: clean-rooms · 2026-10-10 · Documentation low

File: clean-rooms/latest/userguide/analysis-rules-custom.md

Summary

In the custom analysis rule walkthrough for AWS Clean Rooms, the step describing literal comparison controls now notes that every column used in a literal comparison must be in the allowlist 'except for comparisons against always-allowed values', and links to the 'Literal values that are always allowed' section of the custom comparison controls page.

Security assessment

The changed line documents an exception to the allowlist requirement for literal comparison controls, pointing readers to the list of always-allowed literal values. This is privacy/data-egress guidance for Clean Rooms (i.e., documenting which comparison values are permitted by design), so it adds security-relevant privacy documentation but shows no evidence of addressing a specific vulnerability or incident.

Evidence

+  4. Company A adds a custom analysis rule to the `impressions` configured table with a minimum aggregation threshold so that every returned row represents at least 100 distinct users. Company A sets `user_id` as the identity column and overrides the threshold to 5 for the lower-sensitivity `campaign_id` output column. Company A also allows literal comparison on `campaign_id` and `event_date` so that Company B can scope a query to one campaign and a date range. Every column that appears in a literal comparison must be in the allowlist, except for comparisons against always-allowed values. For more information, see [Literal values that are always allowed](./custom-comparison-controls.html#custom-literal-comparison-always-allowed). Finally, Company A adds Company B's account to the allowed analysis providers control so Company B can run queries without per-template review.

Diff

diff --git a/clean-rooms/latest/userguide/analysis-rules-custom.md b/clean-rooms/latest/userguide/analysis-rules-custom.md
index 158467a..f301bc8 100644
--- a//clean-rooms/latest/userguide/analysis-rules-custom.md
+++ b//clean-rooms/latest/userguide/analysis-rules-custom.md
@@ -264 +264 @@ To create a collaboration and run a custom analysis, the companies do the follow
-  4. Company A adds a custom analysis rule to the `impressions` configured table with a minimum aggregation threshold so that every returned row represents at least 100 distinct users. Company A sets `user_id` as the identity column and overrides the threshold to 5 for the lower-sensitivity `campaign_id` output column. Company A also allows literal comparison on `campaign_id` and `event_date` so that Company B can scope a query to one campaign and a date range. Every column that appears in a literal comparison must be in the allowlist. Finally, Company A adds Company B's account to the allowed analysis providers control so Company B can run queries without per-template review.
+  4. Company A adds a custom analysis rule to the `impressions` configured table with a minimum aggregation threshold so that every returned row represents at least 100 distinct users. Company A sets `user_id` as the identity column and overrides the threshold to 5 for the lower-sensitivity `campaign_id` output column. Company A also allows literal comparison on `campaign_id` and `event_date` so that Company B can scope a query to one campaign and a date range. Every column that appears in a literal comparison must be in the allowlist, except for comparisons against always-allowed values. For more information, see [Literal values that are always allowed](./custom-comparison-controls.html#custom-literal-comparison-always-allowed). Finally, Company A adds Company B's account to the allowed analysis providers control so Company B can run queries without per-template review.