The Musk-Altman Trial Matters — But Not for the Reasons Either Side Wants You to Think

The nine-person jury was seated yesterday in an Oakland federal courtroom for what might be the most consequential civil trial in the history of the technology industry. Elon Musk, who contributed roughly $38 million to seed OpenAI as a nonprofit in 2015, is suing Sam Altman, Greg Brockman, and Microsoft for breach of charitable trust and unjust enrichment. He wants up to $150 billion in damages redirected to OpenAI's nonprofit arm¹¹, the removal of Altman and Brockman from leadership, and the unwinding of OpenAI's for-profit conversion. Opening arguments begin Tuesday.

I want to be direct about something: I think Musk will probably lose this case. His legal theories face real problems. As Northwestern law professor Jill Horwitz told MIT Technology Review², the idea that a former donor can sue for breach of charitable trust is "pretty puzzling" when attorneys general already have that authority and already exercised it. The California and Delaware AGs both signed off on OpenAI's restructuring last October³. Musk himself pushed for a for-profit structure as early as 2017 — he just wanted to run it. Polymarket gives him 32% odds⁵. And nonprofit law expert Sam Brunson of Loyola University told Fortune⁶ that donors almost never have standing to sue charities that change direction.

But here's the thing I keep coming back to: whether or not Musk wins the legal case, the trial is producing something the actual regulatory process failed to deliver — genuine public transparency into how OpenAI was built and governed.

The document trail tells a story the attorneys general never forced into the open. The California AG's formal review of OpenAI's conversion began in 2024 and concluded with an agreement in October 2025. During that entire process, the public saw essentially nothing — no founding documents, no internal board communications, no contemporaneous records of who promised what to whom. What we got was an MOU between OpenAI and the California AG¹² laying out structural conditions — but no window into the decisions that got us here.

The litigation, by contrast, has surfaced what The Washington Post described as⁴ "hundreds of court filings" containing emails, texts, and private diary entries from OpenAI's founders and leadership. Greg Brockman's 2017 journal entry — where he wrote something the judge herself quoted² in her January order sending the case to trial — is now part of the public record. So are emails showing Musk told OpenAI leadership in 2018 that the organization "needed billions per year immediately" or should be abandoned. So is a 2023 exchange⁷ where Altman called Musk his "hero" and begged him to stop publicly attacking OpenAI. This isn't curated PR. It's the raw internal record of how an institution that now controls technology affecting nearly a billion weekly users actually made its foundational decisions.

Now, I take the counter-argument seriously: Musk's legal team selected these documents to damage a competitor, not to advance sound governance analysis. The timing of the lawsuit — filed, dismissed, then refiled as xAI launched — tracks competitive pressure, not principle. That's a real concern. A competitor directing what enters the public record, and in what framing, is categorically different from an independent regulator doing so.

But the counter-argument only works if you believe the regulatory process was a functional substitute. I don't think the evidence supports that.

The restructuring the AGs approved has real problems that their review didn't surface publicly. Consider what the "watchdog" arrangement actually looks like. The OpenAI Foundation holds a 26% stake in OpenAI Group PBC, valued at approximately $130 billion⁸. It is technically the controlling entity — it appoints and can remove board members. Sounds strong on paper. But as of April 2026⁸, all but one of the Foundation's board members are also board members of the for-profit. The Foundation is not a private foundation subject to IRS minimum distribution requirements, so it has no obligation to distribute any of its $130 billion in assets⁹. As the Chronicle of Philanthropy reported, a coalition of nonprofit leaders warned that this structure means "the nonprofit structure is just a temporary convenience" — "a way to attract talent and trust, and then, if you become valuable, you can cash out."

The Blue Cross of California precedent¹⁰ is instructive here. When Blue Cross converted in the 1990s, it transferred more than $3 billion in stock to two independent foundations — entities with separate boards and genuine operational independence from the for-profit. Eyes on OpenAI, a coalition of over 60 California nonprofits¹⁰, spent more than a year arguing that the AGs should require the same thing: an independent charitable entity with its own governance, not one that shares its entire board with the company it's supposed to oversee. The AGs did not require this.

This is the gap the trial illuminates, even if accidentally. The regulatory process produced an agreement. The litigation is producing the factual record needed to evaluate whether that agreement was adequate.

Musk's remedy shift is telling, and it cuts in a more complicated direction than either side admits. In January 2026, Musk's lawyers sought up to $134 billion in personal damages. By April, he amended his complaint¹¹ to redirect any damages to OpenAI's charitable arm instead. This is savvy litigation strategy — it blunts the argument that the suit is purely self-serving. But it also means Musk is now, at least formally, seeking a charitable trust remedy rather than a competitive one. His request to remove Altman and Brockman from leadership is harder to frame as purely philanthropic, and I think it reveals the competitive core of the suit. But the damages redirect does move the requested outcome closer to what genuine charitable trust enforcement would look like.

The trial will run for roughly four weeks, with testimony from Musk, Altman, Brockman, and Microsoft CEO Satya Nadella¹. The jury is advisory only — Judge Gonzalez Rogers will make the final determination on liability and remedies herself. She's the same judge who once referred Apple to federal prosecutors⁵ for violating a court order. She is not someone who defers to corporate interests.

So where does this leave us? I think the trial's real significance is neither what Musk wants (competitive destruction of OpenAI) nor what OpenAI claims (a nuisance suit by a jealous ex-founder). The significance is precedential and informational. The $130 billion OpenAI Foundation — the largest nonprofit in America by assets, with shared board members, no payout requirements, and an explicit mission to support the commercial success of the for-profit it ostensibly governs — is a structure that will be scrutinized by every future AI lab, every state AG, and every nonprofit watchdog in the country. The trial is generating the documentary record that makes that scrutiny possible.

The specific indicator I'd watch is this: if the jury's advisory verdict finds liability on the breach of charitable trust claim, even if Judge Gonzalez Rogers ultimately shapes the remedy differently than Musk requests, it will give the California AG leverage to reopen or add conditions to the October 2025 agreement. That, not the damages figure, is the outcome that would reshape AI governance. And if Musk loses entirely, watch for whether the disclosed documents feed Congressional inquiries or state-level legislation on nonprofit-to-profit conversions in the AI sector. The documents are now public. They don't go away when the trial ends.