Three fired OpenAI safety researchers publish an open letter disputing the company’s reasons for dismissing them

Three safety researchers whom OpenAI dismissed last week, Jasmine Wang, Tomek Korbak and Mikita Balesni, published an open letter on Thursday 8 October 2026 disputing the company’s reasons for firing them.^[1][2] The letter is addressed to OpenAI’s Safety and Security Committee, Safety Advisory Group and Mission Advisory Council, and says that “AI is not a normal technology, and OpenAI is not a normal company”.^[1][2] It argues that the freedom to work closely with outside experts “without fear” is “itself an essential safety mechanism”, and that the firings are “chilling the open culture OpenAI has prized in the past”.^[1][2]

OpenAI told AFP, in a statement the agency reported on 2 October, that it had “parted ways with three individuals”, adding: “Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work.”^[5][3] AFP said OpenAI did not confirm the identities, which the Wall Street Journal reported.^[5] OpenAI told TechCrunch the three were dismissed after an investigation found a “pattern of misconduct” that went beyond sharing information with an outside AI evaluation group, and did not say which policies were involved.^[2] OpenAI also told CNN the firings were not for raising safety concerns and that the conduct fell outside legally protected disclosures.^[3]

In the letter, the researchers say they were not the source of a leak to The Information about “supposed new, less monitorable architectures”, and that they do not believe they engaged with external parties outside the mandates of their jobs.^[1] They say the investigation into the Hugging Face incident, in which OpenAI agents breached outside systems during testing, was “without precedent” and that internal policies “were being developed in real time”.^[1][2][3] Korbak was OpenAI’s technical point of contact for the evaluation group METR in that investigation.^[1][6] CNN reported that Korbak said on X he was told he was fired because of how he communicated with METR, and that Balesni said he was told he was “speaking too much to third party safety organizations” and denied leaking intellectual property.^[3]

On Wang, the letter says her access to an executive’s email was delegated for recruiting, that she asked for it to be removed, that IT did not remove it, and that she reported accidentally clicking a sensitive email within minutes.^[1] Wang wrote on X that “The reasons that we were provided for our terminations are simply not adding up”.^[3] The letter also says the researchers did not tell the media of their firings.^[1]

An OpenAI research leader, in an internal memo the company gave to TechCrunch and CNN, wrote: “I want to be very clear that these decisions were not about raising safety concerns or speaking out.”^[2][3] According to CNN, the research leader said they “strongly agree” with the three on the importance of working with outside safety groups, and TechCrunch reported that the memo agrees with the researchers’ recommendations.^[2][3] The letter asks OpenAI to keep its public commitments to embed third-party safety auditors, to preserve the monitorability of frontier models, and to reaffirm an open culture for raising concerns, and says the firings could be used to “justify ending OpenAI’s work with METR”.^[1][4] TechCrunch said OpenAI had not formally responded to the letter.^[2]

Silicon UK reported that METR testified about the Hugging Face incident before the US Senate Homeland Security and Governmental Affairs Committee on 30 September.^[6] Ynet reported that, as of its report, there was no indication that the information allegedly shared was connected to that investigation or that METR or Redwood Research received it.^[7]