Efforts by a group of OpenAI executives and investors to reinstate Sam Altman to his role as chief executive officer reached an impasse over the makeup and role of the board, according to people familiar with the negotiations. Resolution could come quickly, though talks are fluid and ongoing.

Altman, who was fired Friday, is open to returning but wants to see governance changes, including the removal of existing board members, said the people, who asked not to be identified because the negotiations are private. He’s also seeking a statement absolving him of wrongdoing, they said.

After facing intense outrage over the ouster, the board initially agreed in principle to step down, but have so far refused to officially do so. The directors have been vetting candidates for new directors.

Key leaders within OpenAI are also pushing for the board to resign and bring back Altman, one of the people said. The list includes Interim CEO Mira Murati, Chief Strategy Officer Jason Kwon and Chief Operating Officer Brad Lightcap, the person said.

The campaign to reinstate Altman got underway soon after his firing Friday, when the board members led by OpenAI Chief Scientist Ilya Sutskever dismissed Altman and stripped him of his director role, saying "he was not consistently candid in his communications with the board." Hours later, President Greg Brockman, who had also been kicked off the board, resigned in protest of the day’s developments.

One longstanding issue that has divided the company was Altman’s drive to turn OpenAI, which began as a nonprofit organization, into a successful business — and how quickly he wanted the company to crank out products and sign up customers. That ran headlong into board member concerns over the safety of artificial intelligence tools capable of generating text, images and even computer code with minimal prompting.

Until Friday, the company’s board consisted of: Altman, Brockman and Sutskever, as well as Quora Inc. CEO Adam D’Angelo, tech entrepreneur Tasha McCauley and Helen Toner, director of strategy at Georgetown’s Center for Security and Emerging Technology.

At midday Sunday, hints began emerging that a resolution to negotiations may be nearing. Altman and Brockman showed up at OpenAI’s San Francisco headquarters, a person with knowledge of the matter said. On the social site X, Altman later posted an image of himself at the offices wearing a guest badge, with a caption that read, “first and last time I ever wear one of these.”

A deadline has been set for 5 PM San Francisco time for the board to accede to the demands of Altman and his supporters. In order to bring him back as CEO, the board may have to issue a statement absolving him of wrongdoing, according to a person close to the negotiations. However, giving in could leave board members vulnerable from a legal perspective, this person added.

Adding to Altman’s leverage is that he has the loyalty of several employees, people close to him and the company have said. Chief among them are top research scientists seen as the mainstay of OpenAI’s intellectual property and contributors to the large language models at the heart of its groundbreaking ChatGPT chatbot, one of the people said.

Playing a central role in negotiations amind executives, investors and the board is Microsoft Corp. CEO Satya Nadella, some of the people said. Microsoft is OpenAI’s biggest investor, with $13 billion invested.

Bret Taylor, the former co-CEO of Salesforce Inc., will be on the new board, several people said. Another possible addition is an executive from Redmond, Washington-based Microsoft — but it’s unclear whether the software giant can or will take a board seat due to regulatory concerns.

In a memo to staff Saturday, Lightcap said the decision to fire the CEO “was not made in response to malfeasance” or the company’s financial or safety practices.

Altman’s ousting “took us all by surprise,” Lightcap wrote, adding that “we have had multiple conversations with the board to try to better understand the reasons and process behind their decision.”