OpenAI's board of directors fired CEO Sam Altman on November 17, 2023, stating in a brief announcement that he had not been 'consistently candid' with the board, without detailing specific allegations. President and co-founder Greg Brockman resigned in protest within hours, and the two joined Microsoft the following day in a newly created role, with Microsoft CEO Satya Nadella publicly offering to hire any OpenAI employees who wished to follow.
The mechanism behind the crisis was structural: OpenAI's nonprofit board, whose mission was to ensure artificial general intelligence 'benefits all of humanity,' retained legal authority over the for-profit subsidiary that Microsoft had invested roughly $13 billion into and that hundreds of employees held equity in — an arrangement designed in 2019 to subordinate commercial incentives to the nonprofit mission but which had never been tested against a business worth tens of billions of dollars until that week.
Roughly 700 of OpenAI's 770 employees signed an open letter threatening to resign and join Altman at Microsoft unless the board reinstated him and resigned itself, including, notably, Ilya Sutskever, OpenAI's chief scientist and a board member who had voted for Altman's removal and then publicly stated regret over his participation. The near-total employee revolt, combined with Microsoft's public willingness to absorb the entire workforce, made the board's position untenable within days regardless of its underlying rationale.
Altman was reinstated as CEO on November 21, 2023, and the board was reconstituted with new members including former Salesforce co-CEO Bret Taylor and former Treasury Secretary Larry Summers, while Sutskever stepped back from the board and, months later, left the company entirely to found a new AI safety-focused startup. Microsoft gained a nonvoting observer seat on the reformed board, formalizing an influence it had exercised informally throughout the crisis without ever being named the cause.
Coverage during the five days focused heavily on speculation about the board's undisclosed reasons — rumored safety disagreements over a supposedly dangerous research breakthrough gained particular traction online despite no confirmation ever surfacing — over the more mundane and ultimately accurate explanation involving a breakdown in trust and communication between Altman and several board members, including over the board's oversight of a separate entity Altman ran.
Investors watched the episode closely because it tested, in real time, whether a nonprofit board's veto power over a company valued in the tens of billions could actually be exercised without destroying the underlying business — Microsoft's stock dipped only modestly during the crisis, suggesting markets quickly priced in an expectation that Altman's return, not the board's original decision, would prevail.
An independent review commissioned by OpenAI's board and completed in 2024 concluded Altman's conduct did not mandate removal, closing the loop on an episode that had, for five days, threatened to fracture one of the most valuable private technology companies in the world. The episode also accelerated OpenAI's institutional drift toward a more conventional corporate structure, culminating in 2024–2025 restructuring efforts to convert more of the company into a standard for-profit entity, reducing the nonprofit board's practical control.
Sam Altman's brief firing and reinstatement exposed governance tension between nonprofit safety mission and commercial scale. Employees threatened exodus; Microsoft's shadow loomed as both partner and plan B. Cap tables and board structures became front-page AI safety topics.
The crisis educated investors that AI labs' soft power rests on talent loyalty as much as model weights. Subsequent restructuring debates — for-profit conversions, independent committees — inherited that week as precedent. Governance is now part of the product.
Nonprofit-over-for-profit structures looked clever until talent voted with resignation letters. Investors demanded clearer control rights; safety advocates demanded clearer independence. Lab governance is now priced as operational risk beside model evals.
Employees discovered they could veto governance with collective resignation threats. That power is now priced into every lab’s culture. Safety versus speed arguments moved from essays into org charts.
The crisis now functions as the reference case whenever AI governance design is debated: it demonstrated concretely that mission-oriented nonprofit oversight of a commercially explosive AI subsidiary could be overridden within days by employee and investor leverage once the underlying business became large enough, a lesson other AI labs structured with public-benefit or capped-profit models have cited directly when designing their own governance safeguards.
Century Signals note: Contemporaneous investigative reporting on the OpenAI board actions; company and Microsoft statements; later governance analyses. Editorial judgment about what still structures the present — not a comprehensive history.