Researchers say OpenAI revoked their access to limited cyber program
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI revoked access for multiple security researchers to its Trusted Access for Cyber (TAC) program, with affected users seeing messages that their identity "could not be verified" or their account "is ineligible at this time"
- OpenAI confirmed the issue was an error, emailing one researcher that their Daybreak Blue access was revoked "due to a technical issue affecting a limited number of users" and asking them to reapply and complete verification again
- The Daybreak Blue tier, launched August 10, grants vetted individual researchers access to "frontier general-purpose models, including GPT-5.6 Sol" with safeguards tailored to authorized defensive security work
- OpenAI's higher Daybreak Red tier provides models built specifically for cybersecurity research, supporting "authorized vulnerability research, exploit validation, and security testing" for vetted users
- All five affected researchers TechCrunch spoke to live outside the U.S. and Europe, suggesting the revocations may be geographically limited rather than random
- Anthropic runs a comparable program called the Cyber Verification Program (CVP), and both companies have drawn complaints from defensive and offensive researchers who say guardrails block legitimate security work
Why it matters: Security researchers outside the U.S. and Europe lost access to OpenAI's vetted cyber models due to a technical error just weeks after the Daybreak Blue tier launched August 10, forcing re-verification that could delay vulnerability research and push defenders toward Anthropic's competing CVP program.
Ask SkimNews
