Talks at OpenAI to bring back Sam Altman, the artificial intelligence start-up’s recently ousted chief executive, continued into Sunday evening but there were disagreements over the makeup of the company’s board of directors, according to two people familiar with the discussions.
Mr. Altman, 38, spent the weekend waging a pressure campaign on the start-up’s four-person board of directors who ousted him on Friday afternoon, three people familiar with the matter said. The result was a groundswell of support from investors, employees and OpenAI executives.
Mr. Altman was at the OpenAI headquarters on Sunday afternoon. He posted a photo of himself on X, formerly Twitter, wearing a guest identification badge and wrote, “first and last time i ever wear one of these.”
The negotiations included a look at how the company’s board of directors might be reshaped if Mr. Altman returns as chief executive, two of the people said. Members of the board have not yet agreed to what a restructured board of directors might look like — nor is Mr. Altman’s reinstatement an inevitability, two of the people said.
Mr. Altman’s return would like likely include major changes to the board and the company is currently evaluating candidates, according to two people familiar with the situation. Over time, the board’s structure could change to further reflect OpenAI’s evolution from a research start-up to a commercial entity with billions in revenue and enormous ambitions.
But Mr. Altman and the investors still want to retain voices from people who are concerned about the safety of A.I. along with other board members with expertise in technology and commercializing those products, the people said.
OpenAI’s unusual governance structure, which caps its profits, could change, while retaining the checks and balances the company built on the safety of its technology, the people said. The company’s priority is to get a viable board in place by Monday morning, the people said.
Down the line, Mr. Altman and his supporters want to have the company adopt a more traditional structure to reflect its growth into a business valued at $80 billion with many customers and constituents, the people said. They said the events since Friday had been a wakeup call that OpenAI’s structure no longer worked.
There has been a whirlwind of activity since Mr. Altman was forced out of OpenAI, a company he helped found eight years ago that has become one of the most closely watched companies in tech thanks to its popular ChatGPT chatbot.
On Sunday afternoon, Will Hurd, a former OpenAI board member and a former Republican congressman from Texas, was standing outside the company’s headquarters in San Francisco’s Mission district waiting for a ride to the airport after spending two days digging into the details of Mr. Altman’s dismissal.
Mr. Hurd said that a representative of the company had called him on Friday morning, before Mr. Altman’s removal, and asked for his help in navigating the leadership upheaval. Mr. Hurd traveled from Texas to San Francisco on Saturday.
“The industry is important, the company is important,” Mr. Hurd said. “This is the future. How do we make sure there’s a level of trust and transparency? All things we want from models, we want from governance.”
OpenAI declined to comment.
OpenAI’s board of directors has an atypical structure. The organization started as a nonprofit before transforming itself into a for-profit company and bringing on Microsoft as its biggest investor. The for-profit company still answers to the nonprofit board. As a result, the company’s investors have had no official say in what happens to the start-up or who leads it.
Before Mr. Altman was forced out, OpenAI had six board members, including Mr. Altman and Greg Brockman, a company co-founder and board chairman who quit on Friday in solidarity with Mr. Altman.
The other board members are Ilya Sutskever, an OpenAI co-founder; Adam D’Angelo, the chief executive of Quora, the question-and-answer site; Helen Toner, a director of strategy at Georgetown’s Center for Security and Emerging Technology; and Tasha McCauley, an entrepreneur and computer scientist.
Since Friday, people close to the company have been trying to learn why the board dismissed Mr. Altman. Brad Lightcap, the company’s chief operating officer, said in a note to staff on Saturday that there was no “malfeasance” involved.
OpenAI’s board did not tell investors, who have poured billions into the company, what Mr. Altman did wrong. Upon learning of news of his ouster, many investors began a push — in private communications and publicly via social media — to get the board to reverse its decision.
Many reached out to Microsoft, because it is OpenAI’s biggest shareholder and had the most leverage with the company.
The stalemate is the latest twist in a series of power struggles at OpenAI. The fight drew attention to a longtime rift in the A.I. community between people who believe A.I. is the biggest business opportunity in a generation and others who worry that moving too fast could be dangerous.
The company was recently in talks to raise a new funding round that would value the company at more than $80 billion. Bloomberg News earlier reported some of the details of the discussions.
Mr. Altman’s potential reinstatement would be a dramatic reversal for OpenAI, which said he had not been “consistently candid” in his discussions with the board when it unceremoniously ousted him.
As deliberations continued on Sunday, executives at OpenAI called in resources. At 12:45 p.m., a deliveryman with a dozen drinks from the Boba Guys chain showed up on a motorbike outside with two bags. Another deliveryman followed later with a half dozen more drinks.
At 6:15 p.m., a food delivery driver pulled up at the rear entrance and jumped out with four bags from McDonald’s. Two employees from OpenAI loaded them on carts and wheeled them inside for what looked to be a long night.
