Musk v. Altman Enters Final Act: Closing Arguments Begin as Jury Prepares to Deliberate

Three weeks of testimony — featuring a sitting billionaire’s outburst, a LinkedIn co-founder’s name-drop, a gag trophy, and the quiet admission that Microsoft’s own chief technology officer once worried about how OpenAI’s donors would react to a billion-dollar partnership — came to an end on Wednesday evening in a San Francisco federal courtroom. On Thursday morning, closing arguments began in what may be the most consequential corporate governance lawsuit in the history of Silicon Valley.

The case is Musk v. OpenAI. The stakes could hardly be higher: Elon Musk is asking Judge Yvonne Gonzalez Rogers to order the unwinding of OpenAI’s for-profit conversion, remove Sam Altman as CEO, and redistribute up to $134 billion to the original nonprofit entity — a sum roughly equivalent to the GDP of Morocco. OpenAI, now valued by investors at $852 billion, has framed the lawsuit as a jealousy-driven attack by a co-founder who tried to seize control of the company and left when he failed.

What the Jury Heard

The testimony phase of the trial produced a remarkable public record of OpenAI’s early years and the breakdown of its relationship with Musk.

Altman took the stand on Tuesday and Wednesday, offering his most detailed public account yet of why Musk departed. According to Altman, Musk in 2018 proposed that he take majority control — roughly 90% — of OpenAI to “save” it from what Musk claimed was a hopeless funding gap. When the board declined, Musk first threatened to fund a competing AI lab and then quietly left. “He tried to kill OpenAI twice,” Altman testified, referring both to the 2018 departure and to subsequent attempts to disrupt the company’s operations through litigation and public pressure campaigns.

Musk’s attorney launched cross-examination with a single, pointed question: “Are you completely trustworthy?” The exchange set the tone for two days of sparring over whether Altman had promised Musk — in emails, text messages, and verbal assurances — that OpenAI would remain a nonprofit in perpetuity. Altman denied ever making such a promise. Musk claims the entire founding ethos of the organization was premised on one.

The final day of testimony yielded a striking anecdote. In what attorneys described as an unusual moment, a small trophy — reportedly given to OpenAI employees as a memento of the company’s early days — was introduced as a piece of evidence, triggering what witnesses described as a visible emotional reaction from Musk in the gallery. Separately, Microsoft CTO Kevin Scott testified that when his company was weighing its first major OpenAI investment, he had been “primarily thinking about Reid Hoffman” — the LinkedIn co-founder and OpenAI donor — and whether Hoffman would view the Microsoft partnership as a betrayal of the nonprofit’s mission.

The Legal Core

At its heart, the case turns on two civil claims.

Breach of charitable trust. Musk argues that OpenAI was founded as a public benefit organization with explicit commitments to develop artificial general intelligence “for the benefit of humanity” rather than shareholders. Converting to a for-profit structure — even a capped-profit one — violates that foundational trust, he contends. His legal team introduced founding documents, including an early pitch deck circulated to donors, that emphasized the nonprofit structure as essential to OpenAI’s mission.

Unjust enrichment. Musk claims that Altman, Brockman, and Microsoft have materially benefited from a charitable enterprise that was built on donations and non-commercial commitments. By converting the organization and allowing insiders to hold equity in the for-profit entity, he argues, they effectively transferred public assets into private hands.

OpenAI’s defense rests on three pillars. First, that no binding commitment to maintain nonprofit status was ever made to Musk or anyone else; charitable governance documents allowed for restructuring. Second, that the for-profit conversion was necessary to raise the capital required to compete in what became an unexpectedly capital-intensive AI race — one that Musk himself understood would require billions of dollars. Third, that Musk’s lawsuit is not driven by genuine concern about charitable mission but by competitive animus: xAI, his own AI company, competes directly with OpenAI.

The Microsoft Angle

Among the most significant testimony to emerge was that of Microsoft executives, who appeared reluctantly and whose emails and internal communications were cited extensively by both sides. The picture that emerged was of a company that grew deeply anxious about its dependence on OpenAI — even as it invested roughly $13 billion into the organization.

Kevin Scott’s testimony that he had worried about Reid Hoffman’s reaction to the Microsoft partnership was particularly notable. It suggested that even Microsoft, OpenAI’s largest backer, understood that the nonprofit structure was load-bearing — that it was not merely a legal formality but a signal to the research community and donors that OpenAI’s incentives were different from those of a conventional startup. Whether that understanding constitutes a legal commitment is precisely what the jury must decide.

What a Verdict Could Mean

A ruling for Musk would send shockwaves through Silicon Valley. It would potentially unwind one of the most valuable corporate structures ever created and establish a precedent that nonprofit origins can constrain for-profit transformations decades into the future. Practically, it would raise existential questions about OpenAI’s ability to continue operating and raising capital.

A ruling for OpenAI would validate the de facto model by which AI companies have navigated the tension between public-interest framing and private capital: make aspirational nonprofit commitments early, restructure when the economics require it, and defend the transformation as necessary rather than opportunistic.

Neither outcome is tidy. Judge Gonzalez Rogers has already indicated that she views the case as involving genuine legal complexity and is unlikely to simply adopt either side’s framing wholesale. A mixed verdict — finding some breach of duty but awarding significantly less than $134 billion — is considered among the more probable outcomes by legal analysts following the case.

Jury deliberations are expected to begin Thursday afternoon or Friday morning, following closing arguments. A decision could come as quickly as days or stretch into next week. The AI industry, which has watched every day of testimony with unusual intensity, is waiting.