On the stand, Elon Musk can’t escape his own tweets

Elon Musk came to a California federal court on Wednesday to argue that Sam Altman and his co-founders “stole a charity.” He left having admitted, under oath, that Tesla is not currently pursuing artificial general intelligence (AGI)— directly contradicting a tweet he’d posted just weeks earlier.

It was that kind of day for Musk.

The lawsuit he filed challenging the structure of OpenAI alleges Altman and the other co-founders tricked him into backing a non-profit, then launched the frontier lab’s for-profit arm and let it come to dominate the organization. 

After an occasionally testy Musk testified for hours, it appears the case may come down to how much of a distinction jurors and Judge Yvonne Gonzalez Rogers make between investors in OpenAI having their potential profit capped or not. 

In Musk’s telling, when he co-founded the lab with Altman, Ilya Sutskever, Greg Brockman and others, he trusted them to build AI for humanity, but over time became suspicious of their motives, and finally concluded that they were “looting the nonprofit.”

OpenAI’s lawyer William Savitt sought to complicate that story during cross-examination, trying to show that Musk had supported a variety of efforts to transition OpenAI toward for-profit status so it could raise the funds necessary to compete with firms like Google, including incorporating the AI lab into Tesla. 

Musk testified that he had discussed converting the company to a for-profit as early as 2016, and that in 2017, he had explored creating a for-profit arm of OpenAI where he would hold the majority of the equity and control the company. When those plans fell apart, he stopped making regular donations to OpenAI, though he continued to pay for its office space until 2020. 

Musk insisted that there was a big difference between investors whose profits are capped and those whose profits are unlimited. The earliest major investments by Microsoft in OpenAI limited the software giant’s profits, but those restrictions have been rolled back over the years. Musk says those changes ultimately led him to bring this lawsuit.

Savitt tried to establish that Musk had been consulted by Altman and Shivon Zillis — his longtime adviser who is also the mother of four of his children — about subsequent efforts to raise money, and did not object. Zillis was also a member of the OpenAI board when it approved some of those transactions. 

That cross-examination extended to Tesla’s AI ambitions. Notably, Musk was asked about Tesla’s efforts to develop competing AI technologies and found himself, not for the first time, on the wrong side of one of his own posts on X. After Musk said that Tesla’s AI work was focused only on self-driving and not AGI (a term for AI systems that can perform any intellectual task that a human can), he was asked about a recent post claiming that “Tesla will be one of the companies to make AGI.” “We are not pursuing AGI right now,” Musk told the court. (Tesla shareholders may want to take note.)

Musk was also asked about a post where he claimed to have invested $100 million in OpenAI, rather than the $38 million that actually changed hands. He argued that his reputation and network made up for the disparity. 

Savitt brought up emails where Musk had backed efforts by Tesla and his brain interface company, Neuralink, to poach employees from OpenAI while he was still on that company’s board. Another conversation focused on his efforts to hire OpenAI leaders when he left the board in 2018, including Andrej Karpathy, who departed OpenAI to lead self-driving work at Tesla. Musk was also asked about a conversation where Zillis suggested Musk recruit Sutskever to Tesla. 

The most consequential thread of the day, though, may have been about harm prevention. Part of Musk’s case rests on the idea that OpenAI transition into a traditional corporation is dangerous to society because it reduces the company’s focus on safety. Savitt, in turn, had Musk admit that all AI companies, including his own, suffer from this risk. 

Judge Gonzalez Rogers halted that line of questioning, but in remarks to the lawyers after testimony concluded made clear it would resume, with limits. When Musk’s lawyers floated questions about ChatGPT’s role in the Tumbler Ridge shooting—an incident earlier this year in Canada in which a person went on a killing spree after extensive conversations with the chatbot—she made clear that she didn’t want to hear about scandals caused by AI models, but that xAI and OpenAI’s approaches to safety were fair game. 

Musk returns Thursday for another round of adversarial questioning. Also expected to testify are his family office manager, Jared Birchall; AI safety expert Stuart Russell; and OpenAI president Greg Brockman. 

Correction: An earlier version of this story misstated details of the Tumbler Ridge shooting due to an editing error. It has been updated.