Several cybersecurity researchers reported this week that OpenAI abruptly revoked their access to a limited-access program designed for security professionals, according to a report from TechCrunch. The company confirmed the issue was caused by a technical error on its end.
The researchers lost access to the Trusted Access for Cyber (TAC) program, a vetted initiative that allows approved users to employ OpenAI's AI models with fewer restrictions for cybersecurity defense work. The complaints surfaced on OpenAI's official support forums and on social media on Wednesday.
OpenAI stated that the revocations were an error and that affected users are being asked to reapply and complete verification to regain access. The TAC program is intended to give security professionals more flexibility to use advanced AI tools for tasks like analyzing malware or detecting vulnerabilities.
The incident highlights the dependencies researchers are forming on specialized, corporate-controlled AI access programs for critical security work.
