A New Yorker report has reopened the long-running debate over Sam Altman's management style and the governance of OpenAI, saying that board members received secret memos and internal documents accusing him of misrepresenting facts and deceiving executives about safety practices. The report says the material was assembled by Ilya Sutskever, then OpenAI's chief scientist, after weeks of private discussions about whether Altman and Greg Brockman were fit to run the company.
According to the article, the memos were built from roughly 70 pages of Slack messages and HR documents, supplemented by explanatory text and cellphone images. The point of that archive was not just to complain about management style. It was to make the case that Altman was not reliably candid with the board, which had a duty to prioritize safety over ordinary corporate incentives. One memo allegedly opened with the line that Sam exhibits a consistent pattern of lying. That is the level of distrust the report says had developed inside the company.
The story matters because OpenAI was founded with unusually high-stakes governance promises. The New Yorker says the company was structured so that its board would protect humanity rather than maximize revenue, and that the CEO would need to be someone of unusual integrity. Sutskever, the report says, worried that the person leading a civilization-altering technology should not be the kind of person who simply tells people what they want to hear. That anxiety is the backdrop to everything that followed.
The report describes the firing itself as abrupt and chaotic. Altman was in Las Vegas attending a Formula 1 race when Sutskever invited him to a video call and read a short statement saying he was no longer an employee of OpenAI. The board said publicly that Altman was removed because he was not consistently candid in his communications. Investors and executives, including Microsoft chief executive Satya Nadella, were said to have learned of the move only moments before it happened.
The magazine also describes the response as a corporate emergency. Altman returned to San Francisco and set up what he called a sort of government-in-exile at his mansion, where advisers, lawyers and allies spent hours a day on calls and in meetings trying to reverse the decision. The report says the crisis was intense enough that Altman was on calls for more than twelve hours a day and that his camp treated the fight for his return like a war room.
That effort ultimately succeeded, but the report does not treat the aftermath as a clean exoneration. Instead, it presents the episode as evidence of a deeper governance problem: if the people entrusted with AI safety no longer trusted the chief executive, what did that say about the company's structure and the pressure to keep one man in charge? The New Yorker's answer is not explicit, but the question runs through the whole piece.
The source is a magazine feature, not a corporate filing or a court record, so it remains a report of allegations, documents and recollections. Still, it is newsworthy because it puts those claims in one place and shows how close OpenAI came to losing its chief executive over doubts about honesty and internal safety protocols.
