Several cybersecurity researchers experienced unexpected access revocation from OpenAI’s Trusted Access for Cyber (TAC) program. The specialized initiative grants vetted professionals access to advanced AI models with relaxed safety guardrails to help them detect vulnerabilities and analyze potential threats more effectively.
OpenAI quickly clarified that the disruption stemmed from a technical glitch rather than policy enforcement, urging affected users to re-submit their verification details. While normal operations are expected to resume shortly, the incident highlighted the sensitivity surrounding specialized AI tools designed for defensive security work.
- Security researchers locked out of OpenAI’s TAC program
- OpenAI attributed the issue to an internal technical error
- Affected users must complete the verification process again
- TAC provides vetted experts access to less restricted AI models
Sources: