The case in Judge Yvonne Gonzalez Rogers’s Oakland courtroom is, by the time it reaches her in mid-May 2026, the narrowest version of a much larger argument. Of the twenty-six claims Elon Musk filed against Sam Altman, Greg Brockman, OpenAI and the Microsoft Corporation in 2024, two remain at trial. The first is unjust enrichment. The second is breach of charitable trust. Together they ask a simple question with a complicated answer: when Musk wired US$38 million to a nonprofit AI lab in 2015 because he believed the lab existed to keep artificial general intelligence out of corporate hands, did the people he wired the money to enter into a trust with him, and if they did, have they violated it.
The legal architecture
Musk v. Altman is being tried in equity, not at law. That distinction matters. In an equity proceeding the jury is advisory only. Judge Gonzalez Rogers heard testimony alongside the nine-person panel, and she will weigh the panel’s findings, but the binding ruling is hers alone. She has signalled she will deliver it in mid-May 2026.
The trial was bifurcated by the court. The current phase determines liability. If the judge finds for Musk on either remaining claim, a second phase opens to determine damages and remedies. Musk is asking for US$150 billion and the unwinding of OpenAI’s October 2025 conversion to a public-benefit corporation. Disgorgement of what Musk calls Microsoft’s “illicit gains” from the partnership with OpenAI is on the menu in the damages phase.
The amount is unusual. The remedy of unwinding a recently completed corporate conversion is more unusual still. American courts have rarely been asked to reverse a charitable-to-commercial restructuring of a private organisation, and never at this scale.
What the founding-document fight looks like
OpenAI’s lead trial attorney William Savitt has been working a single rhetorical line: Musk read the founding documents loosely at the time and is now seeking to recharacterise them in court. During cross-examination on 28 April 2026 and the two days that followed, Savitt walked Musk through page-by-page comparisons of his earlier depositions and his trial testimony. Inconsistencies surfaced. Judge Gonzalez Rogers intervened multiple times to require Musk to answer yes-or-no questions before continuing.
Musk’s counter-line on the stand was direct. He testified that he gave the money on the understanding that OpenAI would remain a nonprofit dedicated to keeping AGI in public hands, that Altman and Brockman knew that understanding, and that the subsequent corporate transformation breached the conditions under which the money flowed. He used the word “fool” to describe his decision to fund the lab. He said xAI, the AI company he founded after leaving the OpenAI board, distils OpenAI’s models, a concession that will likely matter when the court considers Musk’s standing and motive.
Greg Brockman’s testimony on 5 May 2026 rebutted Musk’s account. Brockman described the 2015 founding from his own vantage and recounted secret work he had performed for Tesla, a detail OpenAI’s team used to push back on Musk’s narrative that the founders had not been straight with him. Satya Nadella took the stand on 11 May 2026 and told the court Musk had never raised concerns with him about Microsoft’s investment in OpenAI. The Nadella testimony complicates the disgorgement theory at the centre of the damages phase.
The settlement-text disclosure
In court filings unsealed in early May, OpenAI’s lawyers revealed that on 25 April 2026, two days before opening arguments, Musk texted Brockman to ask whether OpenAI wanted to settle. Brockman suggested both sides drop their claims. Musk replied: “By the end of this week, you and Sam will be the most hated men in America. If you insist, so it will be.”
OpenAI moved to admit the text into evidence on 4 May, arguing it demonstrated motive and bias. Judge Gonzalez Rogers declined to admit it. The substantive point survived the procedural ruling: the text was reported widely in the days that followed, and it has now framed the public understanding of what Musk is doing and why.
Why a UAP and disclosure archive is paying attention
OpenAI was founded on a transparency premise. The 2015 mission statement committed the organisation to publishing its research, sharing its safety findings, and keeping the most capable artificial intelligence out of the hands of any single corporate or state actor. The October 2025 conversion to a public-benefit corporation, the Microsoft partnership, and the closure of access to frontier-model weights are the specific changes Musk’s complaint asks the court to undo.
Whatever the merits of Musk’s standing to bring the case, the legal theory at the centre of it is the same theory that runs underneath every disclosure conversation: institutions that exist to keep important information in public hands have specific obligations to the public, and when those institutions reorganise away from that mission, somebody has to be able to ask the courts whether the reorganisation was lawful. The trial’s outcome will not by itself establish how AI is governed in the United States. It will, for the first time, generate an appellate-grade record on whether a charitable-trust theory can reach a private AI lab.
What to watch for
Judge Gonzalez Rogers’s ruling is the next decision point. The jury’s advisory finding will be public shortly before it. If she finds liability on either remaining claim, the damages phase opens and Microsoft’s role becomes the central question. If she finds for the defendants, Musk’s path is appellate.
The other ruling that matters arrives separately. The California Attorney General and Delaware Attorney General have been reviewing OpenAI’s restructuring under state nonprofit law since 2024. Either office can act independently of the Oakland verdict. Whatever happens in federal court in mid-May, the state-level review is the other shoe.
Related coverage in the Archive
The Archive has tracked the AI-governance thread through OpenAI’s GPT-5.4 release (March 2026), the Stanford AI Index 2026, Tesla’s Optimus V3 reveal (February 2026), and the broader White House and congressional engagement with AGI risk. Musk v. Altman is the first court proceeding to put the governance question on a binding record. The ruling in mid-May 2026 will be added to this entry as it lands.
