OpenAI glitch locks cyber researchers out of Daybreak Blue access
Vetted researchers suddenly lost access to offensive-security AI models—OpenAI says it's an error.
Several cybersecurity researchers say OpenAI suddenly revoked their access to the Trusted Access for Cyber (TAC) program, which provides vetted experts with fewer guardrails on advanced AI models for defensive security work. Affected users reported seeing "identity could not be verified" or "account ineligible" messages when opening ChatGPT's Cyber page. TechCrunch spoke with five researchers who lost access to Daybreak Blue, the latest individual-researcher tier launched August 10, which includes GPT-5.6 Sol with safeguards tailored for authorized defensive work like vulnerability discovery, malware analysis, and incident response.
OpenAI confirmed the revocations were caused by a technical issue affecting a limited number of users, not a policy change. The company emailed affected researchers and posted on X, asking them to reapply and complete verification to restore access. Notably, all researchers TechCrunch interviewed are based outside the U.S. and Europe, hinting the bug may be region-specific. OpenAI also offers the higher-tier Daybreak Red for offensive security research, but it was not mentioned in the error reports. The incident highlights how access controls on AI cybersecurity tools remain fragile, even for vetted defenders.
- OpenAI revoked access to TAC/Daybreak Blue for at least five security researchers due to a 'technical issue'
- Affected users outside the U.S. and Europe saw 'ineligible' errors and were told to re-verify their identity
- Daybreak Blue includes GPT-5.6 Sol for defensive tasks like vulnerability discovery, malware analysis, and incident response
Why It Matters
Access glitches can stall critical vulnerability research, showing how fragile gates on advanced AI cybersecurity tools remain.