Musk v. Altman Opens Today. The $134B Question Is Whether Charitable Trust Survives a Pivot.

Jury selection in Musk v. Altman starts this morning in federal court in Oakland. Judge Yvonne Gonzalez Rogers, the same judge who ran the Epic v. Apple antitrust trial, will seat nine jurors and there will be no alternates. The trial is scheduled to run through May. Elon Musk, OpenAI, Sam Altman, Greg Brockman, and Microsoft are all parties. The remedy Musk seeks is the unwinding of OpenAI’s for-profit conversion and the removal of Altman and Brockman from their roles.

This is the AI case of the year, and it is not really about AI. It is about whether a charitable trust theory can claw back a corporate restructuring after billions of dollars have already moved.

How four claims survive out of twenty-six

When Musk filed in November 2024, he had 26 claims against OpenAI, Altman, and Brockman. By the time the case got to last week, only four were left: unjust enrichment, fraud, constructive fraud, and breach of charitable trust. On Friday, April 24, Judge Gonzalez Rogers granted Musk’s own motion to drop the fraud and constructive fraud claims, a streamlining move his lawyers wanted to clean up the case for the jury.

What goes to trial: unjust enrichment and breach of charitable trust. Two claims, both equitable, both turning on the same theory. OpenAI was founded as a 501(c)(3) in 2015. Musk donated roughly $44 million. He alleges that he gave that money based on representations that OpenAI would remain a nonprofit dedicated to building safe AI for the benefit of humanity. OpenAI then established a for-profit subsidiary in 2019, restructured again in October 2025, and is now valued at over $850 billion. Musk says the restructuring violated the original charitable purpose and that Altman, Brockman, and the OpenAI board are liable.

The damages number Musk has floated is up to $134 billion. The actual remedy he asked the court for is more interesting than the dollar figure: he wants Altman and Brockman removed and OpenAI’s for-profit conversion unwound.

Why this matters beyond OpenAI

Strip away the personalities. The legal question is whether a donor to a 501(c)(3) can use breach of charitable trust to undo a corporate restructuring years after the fact. The answer matters to every founder, donor, and board member operating in the nonprofit-to-for-profit pipeline that has become standard for ambitious mission-driven companies.

Anthropic was founded as a public benefit corporation. Mozilla operates a for-profit subsidiary under a nonprofit foundation. Wikimedia, the Linux Foundation, and dozens of AI safety nonprofits sit somewhere on the spectrum between charitable purpose and commercial activity. If Musk wins on the charitable trust theory, the doctrine that has let OpenAI raise $14 billion in for-profit capital while preserving a nonprofit governance shell becomes legally fragile.

If Musk loses, the case is a green light. Founders and boards can convert mission-driven nonprofits into commercial enterprises, raise capital, dilute the original mission, and survive donor lawsuits as long as the formal nonprofit shell stays in place.

There is no in-between outcome that does not change the playbook for the next decade of AI company formation.

The Microsoft angle

Microsoft is named as a defendant on an aiding-and-abetting theory tied to the breach of charitable trust claim. Microsoft has invested over $13 billion in OpenAI and holds a contractual relationship that gives it deep economic interest in OpenAI’s for-profit success. Musk’s theory is that Microsoft knew about the original charitable mission and helped engineer the restructuring that violated it.

If the jury finds Microsoft liable on aiding and abetting, the precedent reaches every major investor in every nonprofit-adjacent AI company. Diligence files will get longer overnight. Investors will demand legal opinions on charitable trust exposure before signing term sheets. The cost of capital for mission-driven AI companies goes up.

If Microsoft walks, the case becomes a story about two billionaires fighting and the structural questions get punted to the next case.

What the trial actually looks like

The Washington Post reported last week that hundreds of court filings have already exposed text messages, emails, and diary entries from Musk, Altman, Brockman, and other Silicon Valley figures. Discovery has been brutal. Both sides have ammunition.

Musk’s strongest evidence is the founding documents and early communications where OpenAI’s nonprofit mission was repeatedly affirmed. His weakest evidence is the timeline. He left OpenAI’s board in 2018 over a failed effort to merge it with Tesla, then started his own competing AI company in 2023. The defense will argue that his donation was given freely, that the mission shifted in response to market realities he himself participated in, and that his lawsuit is a competitor’s tactic disguised as donor fidelity.

OpenAI’s strongest evidence is the October 2025 recapitalization, where the company restored a structure with the nonprofit holding a controlling stake in the for-profit subsidiary. They will argue the structure cures any prior deviation. Musk’s response will be that the cure came after the harm and that $850 billion of value has already been allocated in ways the original charter never contemplated.

The jury will hear from Musk under oath. Altman has been preparing for that moment for over a year. Whatever the legal merits, the testimony will produce headlines for weeks.

What general counsel should be doing this week

Three things, depending on what kind of company you advise.

If you advise a nonprofit with commercial ambitions: Pull your founding documents and board minutes. If your nonprofit has founder donors who gave more than $1 million based on a stated mission, you have charitable trust exposure. Document any mission shifts in real time, ideally with donor consent. Get specific legal opinions before any restructuring. The defense that “we gave them notice through the IRS Form 990” will not survive a Musk-style donor lawsuit.

If you advise a for-profit considering nonprofit-adjacent structure: Slow down. The nonprofit-to-commercial pipeline that powered AI fundraising in 2022 to 2025 is about to get tested in a federal courtroom in front of a jury. Wait for the verdict before structuring your next vehicle.

If you advise a corporate investor evaluating mission-driven AI deals: Demand a charitable trust risk opinion as part of diligence. Look for prior donor representations, founding documents, and any restructuring history. The Microsoft theory in this case is the canary. If it survives the motion for directed verdict, your diligence cost just went up.

The likely outcome

Legal experts surveyed by Reuters put Musk’s chances of winning at around 40%. The realistic outcome is a mixed verdict. The jury finds for Musk on unjust enrichment, awards a nominal damages figure that gets multiplied by the donation amount and the appreciation since 2015, and declines to unwind the restructuring as a remedy. Both sides claim victory. Both sides appeal. The Ninth Circuit rules in 2027 or 2028.

That is the path of least resistance for everyone except Musk, who wants the structural remedy. And it is the worst outcome for legal certainty in the AI sector, because it leaves the charitable trust theory alive without resolving when it actually triggers.

The cleanest outcomes are the rare ones. A defense verdict ends the theory. A plaintiff’s verdict with structural remedy ends the playbook. The middle ground keeps the lawyers busy and the founders nervous.

What to do now

Tell your AI investment portfolio companies that the trial starts today and to expect headline volatility. If they are mission-driven AI companies, expect customer questions. If they are taking commercial money on nonprofit-adjacent terms, expect investor questions.

Pull any term sheets that reference OpenAI’s structure as a model. Update the language to flag the pending litigation as a risk factor.

Watch the docket. The interesting motions during trial will signal where the judge is leaning. The directed verdict motion at the close of plaintiff’s case is the moment to pay attention. If Judge Gonzalez Rogers grants partial directed verdict against Musk on the structural remedy, the trial outcome narrows immediately. If she denies it, the jury gets to consider unwinding OpenAI, and every AI company structured as a nonprofit-with-subsidiary should call its lawyers.