A four-week federal case over OpenAI’s soul, a diary entry, $150 billion in claimed damages, and the question of whether a nonprofit can become the world’s most valuable AI company. 

The years-long legal battle between Elon Musk and Sam Altman over the future of OpenAI moved from social media to a federal courtroom in Oakland, California on Monday, as jury selection began in the US District Court for the Northern District of California. Opening arguments are expected on Tuesday.

The trial, presided over by US District Judge Yvonne Gonzalez Rogers, is scheduled to run for four weeks, through mid-May, with court held Monday through Thursday. The jury’s verdict will be advisory: the ultimate decision on liability and any remedies rests with Judge Gonzalez Rogers herself.

The case centres on Musk’s claim that he co-founded OpenAI in 2015 alongside Altman, Greg Brockman, and others with the explicit understanding that it would remain a nonprofit organisation dedicated to developing artificial general intelligence for the benefit of humanity, and that Altman and Brockman deceived him when they converted OpenAI into a for-profit structure in 2019, thirteen months after Musk left the OpenAI board. Musk’s lawsuit, filed in August 2024, alleges breach of charitable trust, fraud, and that Microsoft aided and abetted the breach.

He is seeking $150 billion in damages from OpenAI and Microsoft, with proceeds directed to OpenAI’s charitable arm; he is also seeking Altman’s removal from both the for-profit entity and the nonprofit board, and an order requiring OpenAI to revert to nonprofit status.

Judge Gonzalez Rogers accelerated the core claims to trial because she concluded there is an important public interest in their swift resolution.

OpenAI’s counter-narrative is pointed and supported by internal documents. The company argues that Musk was not deceived about the for-profit transition, that he was, in fact, actively involved in discussions about it, but that he wanted OpenAI merged with Tesla and wanted to lead the combined entity himself.

When Altman and Brockman declined, OpenAI contends, Musk chose to leave and launch his own AI lab. The most consequential discovered document is a diary entry by Brockman, written in autumn 2017, reading: “This is the only chance we have to get out from Elon. Is he the ‘glorious leader’ that I would pick?”

That entry will be central to both sides’ case: for Musk, it evidences a conspiracy to exclude him; for OpenAI, it evidences that OpenAI’s leadership had legitimate concerns about Musk’s ambitions for control.

The witness list is a who’s-who of the AI era. Both Musk and Altman are expected to testify in person. Microsoft CEO Satya Nadella is also expected to appear, as is Shivon Zilis, a former OpenAI board member who is also the mother of four of Musk’s children.

OpenAI’s lawyers have indicated they will argue that Zilis funnelled information about OpenAI to Musk during the period he was no longer on the board, a claim that, if established, would put the personal and corporate dimensions of the story in the same frame.

Depending on how the plaintiffs sequence their witnesses, Musk could be called to the stand as early as Tuesday.

The pre-trial rulings have gone significantly against Musk. He initially sought more than $100 billion in damages for himself personally; after rulings narrowed the available claims, he abandoned personal damages and now seeks the $150 billion to be paid to OpenAI’s charitable arm.

The judge has ruled that Musk cannot be questioned during the trial about alleged ketamine use, a ruling that limits one line of attack on his credibility, but has allowed questioning about his attendance at the 2017 Burning Man festival and about his relationship with Zilis.

The trial also carries personal exposure for Musk that goes beyond the legal outcome: last month, a separate jury held him liable for defrauding investors during his $44 billion Twitter acquisition in 2022, and any damaging testimony about his business conduct in this trial will surface during SpaceX’s planned IPO this summer.

The stakes for OpenAI are existential in a specific sense. If Judge Gonzalez Rogers finds for Musk and orders OpenAI to unwind its for-profit conversion, the company’s ability to raise capital, pursue its planned IPO at a potential $1 trillion valuation, and operate as a commercial enterprise would be directly threatened.

OpenAI has nearly one billion weekly active users, is valued at $852 billion in its most recent round, and has just closed a $122 billion funding round. Microsoft holds a 27% stake in the public benefit corporation. All of that structure depends on the for-profit conversion that Musk is asking the court to reverse.

The case is also, as NPR’s Casey Newton observed, a clash between “two enormous personalities” whose public feud has been one of the defining dramas of the AI era, from Musk’s $97.4 billion offer to buy OpenAI in February 2025 (Altman countered by offering to buy X for $9.74 billion) to their competing posts on X before the trial began.

“Can’t wait to start the trial,” Musk posted in January.

“Really excited to get Elon under oath in a few months, Christmas in April!” Altman responded in February.