OpenAI's "Data controls in ChatGPT" help page now has a section on Trusted Access on personal accounts. It says that during Daybreak sessions on a personal account, OpenAI may use content from your interactions to improve its models even when "Improve the model for everyone" is off, to strengthen safeguards against cyber misuse. The setting still applies to activity outside Trusted Access sessions. Daybreak Blue is OpenAI's vetted access tier that gives defenders fewer refusals for security work.

• Not affected (per OpenAI): ChatGPT Business, Enterprise, Edu and Healthcare workspaces, and API Platform data, which are not used for training by default.
• Dates (from user-posted emails, not the help page): several Daybreak Blue users posted an OpenAI email saying the training change starts Nov 29, and that individual users lose API access to Daybreak Blue on Nov 20 while Codex access through ChatGPT sign-in stays.
• Why it matters: Daybreak sessions typically contain unreported vulnerabilities, PoCs, crash dumps and client data. Researchers on X have objected strongly, and some say they will leave the program. Those reactions are opinion.