Sam Altman testifies in Musk v. Altman OpenAI trial

Strip away the charitable-trust framing and Musk v. Altman is, at its core, a fight over who should have owned and steered OpenAI from the start. Altman's most striking testimony on Tuesday wasn't about governance theory — it was the claim that Musk's opening demand at OpenAI's founding was 90 percent of the equity, a number Altman said his co-founder 'threw out' early in negotiations ^[2]. Altman went further: in a 2017 conversation about creating a for-profit arm, Musk allegedly suggested that control of OpenAI should pass to his children if he died, a moment Altman described as 'hair-raising' on the stand ^[3]. Taken together, the two anecdotes reframe the dispute. Musk's public posture is that he was duped into funding a charity that became a Microsoft-backed cash machine; Altman's counter, reinforced in NBC News coverage of his stand-defense, is that Musk wanted majority control, possibly a Tesla merger, and a dynastic succession plan — and walked away in 2018 only when he could not get them ^[5]. The reason this matters legally is that Musk's two surviving claims hinge on the premise that his donations were given on the strength of an enforceable nonprofit promise. If the jury believes Musk was bargaining for control rather than donating to a charity, the breach-of-charitable-trust theory becomes much harder to sustain — which is why Altman's lawyers spent so much of the day, as CNN's trial coverage detailed, talking about 90 percent and children, not compute and capex ^[6].

Musk filed Musk v. Altman with 26 claims. Only two are still standing at trial: breach of charitable trust and unjust enrichment ^[12]. Both rest on the same idea — that Musk's roughly $38 million in contributions (some filings put the figure at $44 million across 2016 through 2020) were conditional on OpenAI remaining a nonprofit, and that the for-profit conversion converted donor capital into private wealth for insiders ^[9]. Musk's expert David Schizer, the former Columbia Law dean, testified that 'a for-profit subsidiary of a nonprofit corporation is perfectly appropriate if designed to advance the nonprofit's mission, but not when used as a vehicle to create wealth and profit for insiders' ^[11]. The remedy ask is enormous: Musk's team is seeking up to $150 billion in disgorgement, with Musk's own expert pegging the figure at up to $109 billion, plus removal of Altman and Brockman and an unwinding of OpenAI Group PBC ^[2][11]. But the procedural mechanics are unusual. The jury sitting in Oakland is advisory only — Judge Yvonne Gonzalez Rogers alone will set the remedy if liability is found ^[11]. And as attorney Vivian Dong noted in trial analysis, 'it would be unprecedented for a court, in a private breach of charitable trust suit, to order the structural changes to OpenAI that Musk is seeking' ^[11]. The most likely realistic ceiling, in other words, is some form of monetary disgorgement to the nonprofit — not the corporate dismemberment Musk has been campaigning for in public. Axios coverage of Altman's testimony framed the structural-change ask as much more about AI-safety messaging than achievable legal remedy ^[4].

Investors watching the trial care about one number more than $150 billion: $1 trillion. That is roughly the valuation OpenAI is targeting in a planned IPO that the restructure into OpenAI Group PBC was explicitly designed to enable ^[2]. Musk's remedy ask — removing Altman and Brockman, unwinding the for-profit — would force OpenAI to restructure again from scratch and almost certainly delay any public offering. The trial is therefore acting as a multi-billion-dollar overhang on a company that, as of March 2026, had already raised $122 billion at an $852 billion valuation ^[10]. Microsoft is exposed too. Satya Nadella spent his day in the witness box on May 11, and his testimony cut two ways. On one hand, he told the court Musk never raised concerns about Microsoft's investment at the time it was made — a useful fact for OpenAI's defense ^[7]. On the other, prior reporting on his testimony surfaced a worry that Microsoft executives had internally projected roughly $92 billion in returns from the OpenAI deal but also feared becoming 'the next IBM' if Microsoft became overly dependent on OpenAI's IP ^[8]. Microsoft is a named defendant accused of aiding and abetting breach of charitable trust, and an adverse verdict could complicate the contractual IP rights underpinning its tens of billions of dollars of commitments. CNBC's daily trial diary ^[1]has tracked exactly this overhang — closing arguments are Thursday, and a decision could land next week, with both Microsoft's IP rights and OpenAI's IPO calendar contingent on whatever Judge Rogers ultimately does.

The market consensus heading into closing arguments is that Musk faces a difficult legal road. Fortune's Jeremy Kahn argued before Altman's testimony that 'it's far from certain Musk will prevail, either legally, or in shifting public opinion' ^[10]. But Kahn also made a more interesting argument: that for Musk, suing OpenAI may have always been the point, not winning the suit. Discovery has forced OpenAI to produce embarrassing internal documents, fueled investor uncertainty around the restructure at exactly the moment OpenAI is preparing to go public, and given Musk a courtroom megaphone to brand his former co-founders as opportunists ^[10]. The competitive context is hard to ignore. Musk now runs xAI, a direct frontier-model competitor, and during Week 1 of the trial he conceded that xAI 'partly' distills OpenAI's models — an admission that puts a different gloss on his charitable-trust crusade ^[9]. The harshest version of this read: a $1 trillion IPO competitor is being slowed by a lawsuit from a rival who is training his own models on its outputs. Even if Judge Rogers ultimately rejects the structural remedy and awards a manageable disgorgement, the discovery record, the unflattering testimony, and the IPO delay will already have repriced OpenAI in the minds of the investors and regulators who matter.