Sam Altman's Ousting and Reinstatement at OpenAI

Key Insight

The crisis at OpenAI began in November 2023 when CEO Sam Altman was abruptly fired by the board during a Google Meet call. OpenAI's chief scientist, Ilya Sutskever, informed Altman that the board had lost faith in his leadership, providing no specific details before appointing Mira Murati as acting CEO and locking Altman out of his systems. This swift removal was widely termed a 'coup' and involved Sutskever, who reportedly feared super-powerful AI, joining three independent board members, enabling the four votes needed to oust Altman. Microsoft, having invested billions and tied its future to OpenAI, was caught off guard, with CEO Satya Nadella reportedly speechless upon learning the news from Kevin Scott, given Altman was seen as central to their AI strategy.

Microsoft considered three primary options: supporting the new board, leveraging their financial commitment to force Altman's reinstatement and a board restructure, or hiring Altman and his team to build an AI unit within Microsoft. The initial public statements from the OpenAI board were vague, citing Altman's lack of consistent candor, which allowed rumors to proliferate, including theories about his various side hustles. However, the core tension was largely understood to stem from concerns about AI safety and the aggressive pace at which Altman was driving the company, clashing with the watchdog role of the board, which Altman himself had established to ensure AI benefited humanity.

Silicon Valley rallied around Altman, criticizing the independent board members, particularly Tasha McCauley and Helen Toner, for a perceived lack of understanding of entrepreneurial innovation. Employee loyalty was overwhelmingly with Altman, with approximately 745 out of 770 OpenAI staff signing a petition threatening to resign if he was not reinstated; even Sutskever publicly expressed regret for his involvement. Faced with the potential collapse of the company and strong pressure, Altman was ultimately reinstated as CEO. The new three-person board saw the removal of McCauley and Toner, replaced by Bret Taylor (chair) and Larry Summers, with Adam D'Angelo remaining, signifying a shift to a board structure focused on business rather than academic concerns about AI safety.