Over five chaotic days that transfixed Silicon Valley and beyond, the world’s leading artificial intelligence company, OpenAI, appeared to be on the verge of imploding in a power struggle.
The maker of ChatGPT, the sensational chatbot, had a mission to safely develop smarter-than-human AI. But that mission looked in jeopardy on Friday when OpenAI’s non-profit board of directors fired Altman, suggesting he had been dishonest in his communications with them. To many spectators, the future of not just AI but also humanity hung in the balance. Finally late on Tuesday night, after three CEO changes and a full-court press by Microsoft, the board gave Altman his old job back.
Altman’s return had been the subject of days of intense negotiations inside the company, outside pressure from OpenAI’s biggest investor Microsoft, and a threat by nearly all of OpenAI’s employees to quit if he were not reinstated.
Neither Altman nor Greg Brockman, the company’s president who was also removed from the board last week, will immediately retake their seats on the decision-making board. Ilya Sutskever, OpenAI’s chief scientist who delivered Altman the news he was being fired, will reportedly step down from the board but remain at the company. The board’s new chair is Bret Taylor, the former co-CEO of Salesforce. Also on the new board are Larry Summers, the former Treasury Secretary, and Quora CEO Adam D’Angelo, the only remaining member of the board that fired Altman. It is unclear whether more members will be added subject to further negotiations. Satya Nadella, the CEO of Microsoft, called the changes to the board a “first essential step on a path to more stable, well-informed, and effective governance.”
OpenAI did not immediately respond to a request for comment.
More From TIME
As part of the deal, OpenAI will reportedly retain its legal structure: a capped-profit company overseen by a non-profit board that has the legal freedom to take decisions that may not align with the interests of investors. The unusual set-up allowed the board to make its shock decision without informing Microsoft, which had invested more than $11 billion into the company. Altman had established this structure in 2019 to help OpenAI raise more capital while mitigating the perceived risks to humanity of corporate control over artificial intelligence.
Here’s a recap of OpenAI’s rollercoaster week.
The crisis at OpenAI began on Friday last week, when Altman joined a video conference call with the board, minus Brockman. Sutskever informed Altman he was being fired and that the news would shortly be announced, according to a tweet written by Brockman. Sutskever informed Brockman shortly after that he was also being removed from the board, but invited him to remain at the company.
OpenAI then published a blog post on its website announcing the firing. “Mr. Altman’s departure follows a deliberative review process by the board, which concluded that he was not consistently candid in his communications with the board, hindering its ability to exercise its responsibilities,” the post said. “The board no longer has confidence in his ability to continue leading OpenAI.” The board said it had appointed chief technology officer Mira Murati as interim CEO.
While speculation began to swirl about what the board meant by “not consistently candid in his communications,” the board declined to share further information about how or why it had come to its decision. Brockman announced on Twitter that he was resigning from the company in protest at the Board’s firing of Altman, and employee sentiment at OpenAI quickly turned against the board, along with a significant portion of the tech-focused chatter on X, formerly Twitter.
Investors, too, were angry. Microsoft was reportedly only informed of the board’s decision to fire Altman one minute before the blog post was published. Nadella quickly began leading efforts to have the board reinstate Altman at the company, backed up by other OpenAI investors Thrive Capital, Khosla Ventures and Tiger Global Management, according to Bloomberg.
By Saturday, as dozens of OpenAI employees met for talks at Altman’s San Francisco mansion, news had emerged that Altman and Brockman were already pitching a new AI company to investors.
As talks to bring back Altman and Brockman continued late into Sunday at OpenAI’s offices in San Francisco’s Mission District, they appeared to hit an impasse. Late Sunday night, the board named former Twitch chief executive Emmett Shear as OpenAI’s new CEO, replacing Murati, an ally of Altman. Then, Nadella announced that Microsoft would hire Altman, Brockman and other OpenAI staff into a new independent AI unit inside Microsoft, with Altman as CEO.
OpenAI employees, who over the weekend had been engaged in a coordinated show of support for Altman and Brockman on X, began to tweet the same phrase suggesting they might leave en-masse: “OpenAI is nothing without its people.” On Monday, more than 700 of the company’s roughly 800 employees signed their name to an open letter calling on the board to resign, and threatening to jump ship to Microsoft. Shortly after that, Microsoft further increased the pressure, announcing it would offer any OpenAI employee that wanted them roles within this unit for similar compensation.
At the time he was fired, Altman was in talks with investors to sell employee shares in OpenAI at a valuation north of $80 billion, triple its value just six months ago. The deal—the future of which is now uncertain—could have made many of the company’s employees and executives, who are routinely offered equity as part of their compensation packages, extremely wealthy.
The crisis continued into Monday—the first day of a week that was meant to be a company-wide vacation for Thanksgiving, to recognize the amount of work OpenAI employees had put in over a blockbuster year. Instead, OpenAI’s offices were the center of frantic activity.
Sutskever, OpenAI’s chief scientist who fired Altman, made a surprise about-face. “I deeply regret my participation in the board's actions,” he tweeted. “I never intended to harm OpenAI. I love everything we've built together and I will do everything I can to reunite the company.” The change of heart reportedly came after a tearful meeting with Brockman’s wife, Anna, and pressure from significant numbers of company staff.
By late Tuesday, the effort to reinstate Altman as OpenAI’s CEO was complete. Part of the deal included Altman agreeing to an internal investigation into the alleged lack of candor with the board that led to his firing, the Information reported.
Employees celebrated the news with a party at OpenAI’s offices on Tuesday night. Brockman, who returns as president, posted a selfie to X of himself smiling, accompanied by dozens of happy OpenAI employees. “We are so back,” the caption said.
More Must-Reads from TIME
- How the Electoral College Actually Works
- Your Vote Is Safe
- Mel Robbins Will Make You Do It
- Why Vinegar Is So Good for You
- The Surprising Health Benefits of Pain
- You Don’t Have to Dread the End of Daylight Saving
- The 20 Best Halloween TV Episodes of All Time
- Meet TIME's Newest Class of Next Generation Leaders
Write to Billy Perrigo at billy.perrigo@time.com