OpenAI Revokes Researchers' Access to Cyber Program, Cites Technical Error

Several security researchers have reported that OpenAI abruptly revoked their access to a limited-access program designed to relax certain restrictions on using its AI tools for cybersecurity research. OpenAI has confirmed the issue was caused by an error. On Wednesday, multiple researchers raised concerns on OpenAI's official support forums and X, stating that their access to the Trusted Access for Cyber (TAC) program had been revoked. When attempting to open ChatGPT's Cyber page, they encountered messages indicating that their identity could not be verified or that their account was "ineligible at this time." TAC is a specialized program through which OpenAI grants vetted researchers access to its most advanced AI models, with fewer cybersecurity guardrails than those available to regular users. Anthropic offers a similar initiative called the Cyber Verification Program (CVP). The purpose of both programs is to provide trusted defenders with robust tools to identify and report bugs and vulnerabilities to companies, aiming to accelerate patch deployment. Simultaneously, they prevent cybercriminals and malicious actors from accessing these models to discover exploitable flaws. To qualify for TAC, cybersecurity researchers must submit identification and undergo vetting by OpenAI. At this stage, it remains unclear why these researchers are losing access or how many individuals are affected. TechCrunch spoke with five researchers who reported this problem. One researcher shared an email from OpenAI stating that access to Daybreak Blue, the latest vetted tier of TAC, was revoked "due to a technical issue affecting a limited number of users." The message emphasized, "This was an issue on our end, and not the user experience we want to deliver." In an OpenAI forum thread, another researcher noted that after contacting official support, the company attributed the access loss to a "recent technical issue." Both communications advised affected users to reapply and complete the verification process again. All researchers interviewed by TechCrunch reside outside the U.S. and Europe, suggesting the revocations may be geographically limited. OpenAI directed TechCrunch to a tweet stating that a "limited set of users' access to Daybreak Blue is no longer active and they will need to re-verify to maintain their access." [Note: The original article appears to be cut off. The response ends here as per the provided content.]

via TechCrunch AI