OpenAI's five-day CEO crisis (November 2023): the firing, the employee uprising, and the rehiring
On Friday afternoon, November 17, 2023, the OpenAI board fired co-founder and CEO Sam Altman. The board statement said Altman had not been “consistently candid in his communications with the board, hindering its ability to exercise its responsibilities.” Within hours the situation escalated. Microsoft — OpenAI's largest commercial partner — offered to hire Altman to lead a new AI research unit. Roughly 95% of OpenAI's 770 employees signed an open letter threatening to resign and follow Altman to Microsoft unless the board reinstated him. Investors and partners pressed publicly. Two interim CEOs (Mira Murati, then Emmett Shear) held the role across the weekend. By Tuesday November 21, Altman was back as CEO, several of the firing-vote board members were out, and a new board had been seated. The whole episode is the highest-profile recent example of how concentrated employee, investor, and partner power can override a formal board decision in a frontier technology company.
- Story: OpenAI board fired CEO Sam Altman November 17, 2023. Within 5 days: Microsoft offered to hire Altman, 700+ of ~770 OpenAI employees threatened to follow, Greg Brockman resigned. Altman reinstated November 22 with new board. Defining corporate governance crisis.
- Why it matters: OpenAI Altman firing is a defining recent corporate governance crisis case — demonstrating unusual governance structures can produce crisis when board interests diverge from stakeholder interests.
- Takeaway: Unusual governance structures can produce crisis when board interests diverge from commercial subsidiary stakeholders.
- Takeaway: Major investor relationships create constraints on board decisions.
- Takeaway: Employee loyalty can be decisive factor in CEO-vs-board conflicts.
OpenAI Altman firing — the four-step story
OpenAI Altman crisis by the numbers
Quick facts
The OpenAI structure that made the firing possible
OpenAI was founded as a non-profit research lab in 2015. In 2019 the organisation added a capped-profit subsidiary (OpenAI LP, later restructured) that could raise outside capital while operating under the governance of the original non-profit board. That structure made the non-profit board legally able to fire the CEO of the for-profit operating company without shareholder approval, because the non-profit had full control over the for-profit subsidiary. The structure had been designed specifically to keep the safety-focused mission supreme over commercial pressure.
By late 2023 the non-profit board consisted of Sam Altman, Greg Brockman (OpenAI president), Ilya Sutskever (chief scientist), Helen Toner (Georgetown CSET), Tasha McCauley (RAND), Adam D'Angelo (Quora CEO), and Reid Hoffman/Will Hurd had recently left. Internal disagreements between Altman and at least some independent directors had been building, reportedly around the pace of capability development, commercial commitments, and Altman's communication with the board.
The five-day timeline
Friday November 17 afternoon: the board announced Altman's firing via blog post, citing a lack of candour with the board. Mira Murati, then CTO, was named interim CEO. Greg Brockman was removed from the board chairmanship and shortly resigned in solidarity with Altman.
Saturday-Sunday: investor pressure mounted. Microsoft, Thrive Capital, Tiger Global, and Sequoia all pressed publicly and privately for Altman's reinstatement. By Sunday evening it was clear that Murati would not stay in the interim role; Emmett Shear (former Twitch CEO) was named as the next interim CEO. Satya Nadella publicly stated that Microsoft would hire Altman and Brockman to run a new in-house AI research unit, effectively forcing the board to choose between reinstating Altman or losing the company to Microsoft.
Monday November 20: an open letter signed by roughly 95 percent of OpenAI's ~770 employees, including Ilya Sutskever (who had voted for the firing on Friday), demanded Altman's reinstatement and the resignation of the firing-vote board members. The letter threatened mass resignation and migration to Microsoft.
Tuesday November 21: Altman was reinstated as CEO. Helen Toner, Tasha McCauley, and others left the board. A new initial board was seated: Bret Taylor (chair, former Salesforce co-CEO), Larry Summers, and Adam D'Angelo (the only continuing board member from the firing vote). Subsequent additions broadened the board further.
What the case shows about governance
The OpenAI structure was specifically designed to keep mission-aligned board members in control of CEO decisions. The structure worked legally — the board did fire Altman, and the firing was within its authority. But the structure did not survive contact with the realities of a frontier technology company that depended on its employees, its largest commercial partner (Microsoft), and its investors. The board's formal authority was overridden by the practical truth that without Altman, the company's humans and capital would have walked.
The subsequent WilmerHale independent review (commissioned by the new board) concluded that the original board had acted within its authority but did not find that Altman's conduct mandated removal. The framing matters: the case is not that the board was wrong to act, but that the structure created an inevitable mismatch between formal governance authority and practical operating authority.
How RGM thinks about governance and key-person dynamics
When clients ask about governance structures intended to preserve mission over commercial pressure, the OpenAI case is the defining recent example of the limits. Three structural lessons. First, formal authority that is not aligned with practical operating power is fragile under stress. The board legally fired Altman; the company nearly disintegrated within 48 hours. Second, employee, investor, and partner concentration creates a parallel power structure that mission-aligned boards cannot easily override. OpenAI's ~770 employees, Microsoft as a strategic partner, and the major capital providers collectively had more practical control than the board did. Third, the cost of an unsuccessful governance action is asymmetric: a board action that succeeds reinforces governance authority; a board action that fails undermines all subsequent governance authority.
The pattern is hard to copy without the specific scaffolding OpenAI had (non-profit parent over for-profit subsidiary, mission language in the structure, founder concentration). Clients trying to build mission-protected governance structures should study the failure modes of this case as carefully as the structural design. The OpenAI structure has since been substantially modified, and a broader corporate restructuring of OpenAI itself was announced in subsequent years to align the governance structure with the operating reality.
Frequently asked questions
Why did the board fire Sam Altman?
The board's public statement cited a lack of candour: Altman was “not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities.” Specific underlying issues have been partially disclosed: the board reportedly learned about the November 2022 ChatGPT release from Twitter rather than from Altman, and there were concerns about Altman's undisclosed ownership of the OpenAI Startup Fund. Helen Toner has spoken publicly about additional concerns. The full board reasoning has not been comprehensively disclosed.
How long was the crisis?
Five days. Altman was fired Friday afternoon November 17, 2023 and reinstated Tuesday November 21, 2023. The intervening weekend was dominated by Microsoft, employee, and investor pressure on the board to reverse course.
Who were the interim CEOs?
Mira Murati, then OpenAI CTO, served as interim CEO from Friday to Sunday. Emmett Shear, former Twitch CEO, was named the next interim CEO Sunday evening but served only until Tuesday when Altman was reinstated. Neither held the role long enough to make material strategic decisions.
What was Microsoft's role?
Microsoft is OpenAI's largest commercial partner and had invested approximately $13 billion in OpenAI through cloud-services credits and direct capital. Satya Nadella publicly stated that Microsoft would hire Altman and Brockman to run a new AI research unit, effectively creating a parallel landing pad for OpenAI's employees if Altman were not reinstated. The threat was credible and accelerated the board's reversal.
How did the board change afterward?
Helen Toner and Tasha McCauley (who had voted to fire Altman) left the board. Adam D'Angelo remained as the only continuing director from the firing vote. New initial board members included Bret Taylor (chair) and Larry Summers. The board was subsequently expanded further to include additional independent directors and to broaden technical and policy expertise.
Has anything structural changed since?
Yes. OpenAI's corporate structure has been the subject of subsequent restructuring discussions and announcements, including proposed shifts toward a more conventional for-profit operating structure that aligns governance authority with operating reality. The independent WilmerHale review concluded the original board had acted within its authority but did not find Altman's conduct mandated removal.
Sources & references
- Removal of Sam Altman from OpenAI (Wikipedia) — Aggregated reference for the five-day timeline, board composition, and key events.
- Former OpenAI board member explains why Sam Altman got fired (CNBC) — CNBC coverage of Helen Toner's subsequent public account of the firing rationale.
- OpenAI's CEO Crisis Pitted Sam Altman Against Ilya Sutskever (Biography.com) — Editorial retrospective on the internal dynamics between Altman, Sutskever, and the board.
- Sam Altman's Firing And Rehiring by OpenAI (Advice Scout) — Industry analysis of the events and governance implications.