San Francisco — Fired OpenAI Safety Researchers Dispute Misconduct

Three safety researchers fired by OpenAI last week published an open letter on Thursday, Oct. 8, 2026, disputing the company's account of their dismissals and warning that the manner of their firing is discouraging current employees from raising safety concerns, according to TechCrunch, which obtained the letter.

The letter is signed by Jasmine Wang, Tomek Korbak, and Mikita Balesni, addressed to OpenAI's Safety and Security Committee, Safety Advisory Group, and Mission Advisory Council, per TechCrunch's Oct. 8 report.

What does the researchers' letter say?

"We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in ways that, until last week, were an integral part of working at OpenAI," the three wrote, according to the letter text TechCrunch published.

The researchers argue that outside collaboration is itself a safety function, not a policy violation. "Those of us who work on safety see risks before anyone else, and we rely on close collaboration with outside experts to work out how to address them," they wrote. "The freedom to do so without fear, and to have well-defined internal procedures that enable this work, is itself an essential safety mechanism."

They describe a shift in company culture away from one that "encouraged workers to raise safety concerns and disagree openly," writing that staff are now "unclear on where they stand" when conduct treated as normal a month earlier became grounds for termination. "Given the significant safety concerns surrounding the development of AI, employees must not be left working in an environment where fear and unclear rules stymie AI safety work and weaken third-party accountability," the letter states.

Why did OpenAI fire the three researchers?

OpenAI dismissed Wang, Korbak, and Balesni last week after accusing them of sharing confidential company information with a third-party AI safety organization outside established procedures, TechCrunch reported. The company said the three violated internal policy by "accessing and handling sensitive company information."

TechCrunch's reporting does not name the third-party organization or specify the information at issue beyond OpenAI's characterization.

What specific allegations do the researchers deny?

The letter denies two distinct claims. First, the three say they were not involved in a leak to The Information concerning less-monitorable architectures in OpenAI's newest models — reporting that described changes making chain-of-thought reasoning harder to monitor, per TechCrunch's account of the letter. Second, they deny engaging with external parties outside the mandates of their jobs, directly contradicting OpenAI's stated rationale for the firings.

The letter does not address every detail of OpenAI's internal findings, and TechCrunch's report does not include a point-by-point rebuttal beyond these two denials.

How has OpenAI responded to the letter?

OpenAI has not issued a formal public response to the open letter itself, according to TechCrunch. The company did share with TechCrunch an internal memo attributed to a research leader that praises the three researchers' prior contributions to AI safety while rejecting any suggestion of retaliation.

"I want to be very clear that these decisions were not about raising safety concerns or speaking out," the memo reads, per TechCrunch. "We have always encouraged that and always will. We do not terminate employees for raising concerns."

The memo does not, in the portion TechCrunch published, rebut the researchers' specific denials regarding the Information leak or their characterization of their external contacts as within job scope.

What does this mean for OpenAI's safety culture going forward?

The dispute turns on two incompatible accounts of the same set of facts: OpenAI frames the firings as enforcement of data-handling policy, while the three researchers frame them as punishment for safety-oriented external engagement that the company previously tolerated. TechCrunch's report presents both versions without resolving which is accurate, and no independent verification of the underlying information-sharing incident appears in the available reporting.

The letter's core empirical claim — that colleagues inside OpenAI are now less willing to speak to outside safety researchers or raise internal concerns — is not independently measured in the source material; it rests on the three authors' own assessment of internal reaction, not on a survey or dataset. OpenAI's memo likewise offers no figures on departures, morale, or changes in external-collaboration volume since the firings.

What is documented: three named individuals were terminated in the same week, both sides have gone on record with conflicting explanations, and the dispute is playing out inside formal OpenAI governance bodies — the Safety and Security Committee, the Safety Advisory Group, and the Mission Advisory Council — rather than only in public statements, per the addressees listed on the letter TechCrunch obtained.

For the full letter text and OpenAI's internal memo, see TechCrunch's Oct. 8 report.

mindfulAI Keyboard — An iPhone keyboard that rewrites a draft in a tone you pick. Four free AI actions a day.