OpenAI has revoked access to its Trusted Access for Cyber (TAC) program for several security researchers, according to reports on OpenAI‘s forums and statements from five researchers who spoke to TechCrunch.

The researchers received emails stating their access to Daybreak Blue, the latest vetted tier granting access to frontier models like GPT-5.

6 Sol with reduced guardrails for defensive security work, was removed due to a “technical issue affecting a limited number of users.

” OpenAI acknowledged the error and asked affected users to reapply and complete verification. Notably, all researchers affected in these reports are based outside the U.S.

and Europe, suggesting the revocation may be region-specific. OpenAI‘s public response, via a tweet, confirmed that a limited set of users lost access and must re-verify.

The incident follows ongoing complaints from both defensive and offensive security researchers about guardrails imposed by OpenAI and Anthropic, arguing these restrictions hinder legitimate vulnerability research and responsible disclosure work.

OpenAI launched Daybreak Blue on August 10 as the entry-level tier for individual researchers, along with a higher tier called Daybreak Red for more specialized offensive security research.

The exact scope of the revocation and its regional or technical causes remain unclear, but the company maintains it was an error on their end.