Musk Takes the Stand for Two Days as OpenAI Trial Wraps Its Explosive First Week

The courtroom in Oakland, California did not disappoint. When the first week of Musk v. OpenAI ended on Thursday, April 30, Elon Musk had occupied the witness stand for nearly two full days, delivered some of the most televised Silicon Valley testimony since the early social media antitrust hearings, and set the stage for what observers expect to be an equally combustible second week when Sam Altman takes the stand.

The stakes remain enormous. Musk is seeking $130 billion in damages and a court order that would require OpenAI to abandon its for-profit corporate conversion and return to a nonprofit structure. He also wants Altman and co-founder Greg Brockman removed from the company’s board. Judge Yvonne Gonzalez Rogers is presiding; a nine-member advisory jury was seated on opening day, April 27, and the judge is expected to deliver her ruling by mid-May.

What Musk Argued on the Stand

Musk’s core narrative, which he began laying out on Day 2 (Tuesday, April 28) and continued through Day 3 (Wednesday, April 29), centers on a founding premise he says was betrayed. In his telling, OpenAI was incorporated as a nonprofit in 2015 precisely to ensure that transformative AI would be developed for humanity’s benefit rather than concentrated in the hands of shareholders. Google, he testified, was the threat he was trying to counter. He funded the organization, he said, because Altman and others agreed that the nonprofit structure was sacrosanct.

“It’s not OK to loot a charity,” Musk said on the stand, a phrase that summarized his legal argument with unusual directness. He cast the conversion of OpenAI from nonprofit to a capped-profit structure, and now a full for-profit public benefit corporation preparing for an IPO, as a violation of the founding promise.

Musk also used his platform to take a broader swipe at Altman’s AI safety credibility. Per Axios, Musk repeatedly positioned himself as the more responsible AI actor — pointing to his founding of xAI and the continued open-source release of Grok models as evidence of his commitment to transparency, in contrast to what he characterized as OpenAI’s shift toward closed, proprietary development after securing Microsoft’s $13 billion investment.

The 2018 Term Sheet: The Trial’s Central Flashpoint

The most heated exchange of the week came on Day 4 (Thursday, April 30), when OpenAI’s lead attorney Will Savitt returned to cross-examine Musk about a 2018 term sheet that is central to the damages claim.

In his original deposition, Musk had made no mention of reading the term sheet in detail. On the stand Wednesday, however, he said he had read “the very first part” and gleaned the high-level details. Savitt played the deposition video in open court, then confronted Musk with the discrepancy directly.

The exchange, which the Oakland courtroom reportedly watched in near-silence, became the week’s sharpest illustration of the credibility questions that will ultimately determine the outcome. Musk’s legal team had presented him as a principled whistleblower; OpenAI’s team has spent the week constructing a counter-narrative in which Musk is a litigant with selective memory and financial motives — his 2023 attempt to acquire OpenAI for $97 billion, and his subsequent founding of a competing AI lab, feature prominently in that narrative.

Russell Cohen, an attorney for Microsoft — a co-defendant in the suit — also questioned Musk on Thursday, pressing on the company’s relationship to OpenAI’s early funding and development.

The Birchall Testimony

On Day 4, OpenAI attorney Bradley R. Wilson questioned Jared Birchall, the head of Musk’s family office, about the structure of donations Musk made to OpenAI through donor-advised funds. Wilson’s questions, and Birchall’s answers, appeared designed to establish that OpenAI leadership believed at the time that donated funds were theirs to direct — a framing that cuts against Musk’s claim that he was deceived about how his contributions would be used.

The Birchall testimony underscored the broader evidentiary picture: OpenAI’s defense is less about denying that the organization changed its structure, and more about arguing that the change was legally permissible and commercially necessary — and that Musk, a sophisticated investor and entrepreneur, understood or should have understood the terms of his involvement.

What Week Two Holds

The jury was dismissed for the long weekend with strict instructions from Judge Gonzalez Rogers not to discuss the case or conduct research. Trial proceedings resume Monday morning.

Sam Altman is expected to take the witness stand during week two — a moment that most observers consider the centerpiece of the case. Altman’s testimony will be the most direct counterpoint to Musk’s account of the founding, and his legal team’s challenge is to present him as a mission-driven CEO who made difficult but principled decisions to ensure OpenAI could compete with Google and Anthropic, rather than as an opportunist who leveraged a charitable structure for personal and corporate enrichment.

Greg Brockman, who has been on an indefinite leave of absence from OpenAI, is also expected to testify.

Why This Case Matters Beyond the Courthouse

Even setting aside the $130 billion damages figure, the trial is producing a running public record of AI’s most consequential startup decisions — and the tensions at the industry’s heart. The competing narratives of safety versus commercialization, open development versus closed proprietary systems, nonprofit mission versus for-profit imperatives, are not just legal claims. They are the central debates of the AI era, playing out under oath in a federal courtroom.

Judge Gonzalez Rogers’s final ruling will carry legal weight, but the trial’s broader significance may be the clarity it forces on questions that the tech industry has largely allowed to remain ambiguous: What does it mean to build AI “for humanity”? Who holds the obligation when a nonprofit pivots? And can the most powerful technology in decades be adequately governed by the organizations that created it?

Those questions don’t resolve in mid-May. But the answers developing in Oakland are shaping the framework in which they will be asked for years to come.