OpenAI Trial Concludes with Final Testimony Involving Musk Outburst and Hoffman Dispute

Three weeks of testimony. Nine jurors. Two remaining legal claims. And a golden jackass trophy that never made it into evidence.

The trial of Musk v. Altman wrapped up on May 14, 2026, in Judge Yvonne Gonzalez Rogers' courtroom in Oakland, California, with closing arguments scheduled for the following day [1]. What began as Elon Musk's 26-claim lawsuit against OpenAI, CEO Sam Altman, and president Greg Brockman has been whittled down to two claims: unjust enrichment and breach of charitable trust [2]. But the case's implications stretch far beyond these two causes of action — they reach into the foundational question of whether the most consequential technology of the century can be built inside a charity, and what happens when the charity decides it cannot.

The Legal Architecture: What Remains and What Was Cut

Musk originally filed suit in February 2024, alleging breach of contract, breach of fiduciary duty, fraud, false advertising, unfair business practices, unjust enrichment, and breach of charitable trust [3]. Judge Gonzalez Rogers trimmed the case significantly before trial, excluding the fraud claims (which Musk himself dropped days before trial began [4]), the false advertising claims, and the breach of fiduciary duty claims.

The two surviving claims center on a core question: did Altman and Brockman's communications with Musk around OpenAI's 2015 founding establish a formal "charitable trust" — a legal arrangement requiring that assets be used exclusively for their stated charitable purpose? [2] If so, OpenAI's conversion to a for-profit entity may have violated that trust.

Musk's legal team initially sought $134 billion in "wrongful gains" [5], later revising the request to ask that up to $150 billion be "disgorged" — returned — to the nonprofit entity [6]. The figure approximates OpenAI's valuation at the time the lawsuit was filed and has grown as the company's valuation has soared.

The jury in this case is advisory only. Judge Gonzalez Rogers will issue the final ruling on both liability and remedies [2]. The advisory jury's recommendation is expected as early as the week of May 18.

The Musk Outburst: A "Jackass" for Safety

The trial's final day of testimony produced one of its most memorable moments, courtesy of OpenAI's chief futurist, Joshua Achiam [1].

Achiam testified about a 2018 companywide meeting where Musk addressed roughly 50 to 60 employees about his planned departure from OpenAI. Musk told the group he wanted to "build it very fast, because he was very worried that someone else, if they got it, would do the wrong thing with it" [1]. When Achiam challenged Musk on whether that approach was safe, calling it "unsafe and reckless," Musk responded by calling Achiam a "jackass" [1].

OpenAI's lawyers attempted to introduce a physical piece of evidence corroborating the story: a small golden jackass trophy, inscribed with "never stop being a jackass for safety," which colleagues Dario Amodei (now CEO of Anthropic) and David Luan had given Achiam as a thank-you for standing up to Musk [1]. Judge Gonzalez Rogers declined to admit the trophy as evidence, and it was never shown to the jury [1].

The exchange, while colorful, carried substantive weight for OpenAI's defense. It suggested that Musk's approach to AI development in 2018 — speed above safety — contradicted the charitable mission he now claims to be defending.

Reid Hoffman and the Donor Question

Reid Hoffman, the LinkedIn co-founder and early OpenAI donor, featured prominently in the trial's final stretch through testimony by Microsoft CTO Kevin Scott [1].

Scott testified about a 2018 email in which he raised concerns about how OpenAI's early nonprofit donors, including Hoffman, would react to a commercial partnership with Microsoft. Scott wrote that he doubted donors would appreciate OpenAI using their seed money to "go build a for-profit thing" [1]. The email underscored a tension at the heart of the case: whether the funds Musk and other early backers contributed carried implicit or explicit restrictions on how they could be used.

Hoffman was not a defendant in the trial. His role was that of a figure referenced in testimony and documentary evidence — a data point in the argument over whether OpenAI's leadership understood that its nonprofit donors expected their contributions to remain in a nonprofit structure [7].

Altman's Defense: "He Didn't Steal a Charity"

Sam Altman took the stand on May 12, delivering what NPR described as an effort "to fend off Elon Musk's accusations he 'stole a charity'" [8]. Altman's central claim to the jury was direct: "He didn't steal a charity, but Elon Musk abandoned one" [9].

Altman testified that he never promised Musk that OpenAI would remain a nonprofit in perpetuity [10]. He said Musk wanted "total control" of any for-profit entity OpenAI might create, with a promise to reduce that control over time — a proposal the other co-founders rejected [8]. Altman further testified that Musk had proposed merging OpenAI with Tesla, which the founders also declined [9].

On cross-examination, Musk's lawyer pressed Altman on his trustworthiness, asking: "Are you completely trustworthy?" [11]. Musk's legal team also introduced testimony from multiple former associates who had questioned Altman's honesty, including Anthropic CEO Dario Amodei, who allegedly accused Altman of misrepresenting the terms of an investment [12].

Altman described OpenAI after Musk's departure in 2018 as having been "left for dead," arguing that the shift to a capped-profit model was a survival decision, not a betrayal [9].

Brockman, Zilis, and the Control Narrative

Greg Brockman, OpenAI's co-founder and president, testified during the second week that Musk had wanted the company to create a for-profit entity — contradicting Musk's position that the for-profit conversion betrayed the founding agreement [13]. Brockman also revealed that he had done secret work for Tesla while at OpenAI, a detail that complicated the narrative of clean institutional boundaries [13].

Perhaps the most revealing testimony came from Shivon Zilis, a former OpenAI board member and mother of several of Musk's children. Zilis testified that Musk had tried to recruit Altman to Tesla's board and had also attempted to "poach" Altman from OpenAI [14][15]. Her testimony suggested that Musk's relationship with OpenAI's leadership was more entangled — and more personally competitive — than his legal filings implied.

The Microsoft Factor

Microsoft CEO Satya Nadella testified for roughly two and a half hours, telling the court that Musk never contacted him with concerns that Microsoft's investments in OpenAI violated any special terms or commitments [7]. Internal Microsoft documents revealed during testimony showed the company had projected a $92 billion return on its cumulative $13 billion investment in OpenAI [16].

An April 2022 email from Nadella surfaced during testimony in which the CEO wrote: "I don't want to be IBM and OpenAI to be Microsoft" — a candid expression of concern about becoming too dependent on a single AI partner [17].

Kevin Scott's testimony also revealed that Microsoft internally worried about overreliance on OpenAI, adding nuance to the portrait of the partnership [17].

The Money at Stake

The financial transformation of OpenAI is central to Musk's claims. The numbers tell a story of exponential growth that made early charitable contributions look, in hindsight, like the most lucrative seed investments in technology history.

Source: Sacra / OpenAI disclosures

Data as of May 1, 2026CSV

OpenAI's valuation has increased from roughly $27 billion in 2020 to $852 billion as of March 2026, when the company closed a $122 billion funding round co-led by SoftBank, Andreessen Horowitz, and other investors [18]. In October 2025, OpenAI formally converted from its nonprofit-controlled capped-profit structure into a Public Benefit Corporation (PBC), a move that made a traditional IPO possible [19]. The original nonprofit was renamed the OpenAI Foundation and retained a 25.8% equity stake in the new PBC [19].

Source: OpenAI / Sacra estimates

Data as of May 1, 2026CSV

Revenue has grown from approximately $30 million in 2022 to an estimated $25 billion annualized in early 2026 [18]. Musk's legal team argued that this growth — built on research funded by nonprofit donations — demonstrates that the mission shift harmed the public interest the charity was supposed to serve [6].

OpenAI counters that these financial results validate the for-profit conversion: a nonprofit structure could never have attracted the capital needed to compete with Google, Microsoft's own internal AI efforts, and well-funded startups [9].

The xAI Problem: Competitive Motive or Principled Stand?

OpenAI's defense rested heavily on undermining Musk's stated altruistic motivations by pointing to his own commercial AI ambitions.

Musk founded xAI in 2023, shortly after OpenAI's ChatGPT sparked a commercial AI boom [20]. During trial, Musk admitted that xAI uses OpenAI's models to train its own chatbot, Grok — a practice known as "distillation" [20]. OpenAI's lawyers argued this admission proved that Musk was suing a competitor while simultaneously free-riding on its technology.

Days before the trial began, Musk contacted Brockman to "gauge interest in settlement" [21]. When Brockman suggested both sides drop their claims, Musk replied: "By the end of this week, you and Sam will be the most hated men in America" [21]. OpenAI's attorneys introduced this exchange as evidence that "Mr. Musk's motivation in pursuing this lawsuit is to attack a competitor and its principals" [21].

During the trial's first week, Musk also flew to China while proceedings were still active, prompting Judge Gonzalez Rogers to warn that he could be called back to court [22].

Legal Precedent: The Wrong Body of Law?

Several legal scholars have questioned whether the case is even being litigated under the correct legal framework.

Jill Horwitz, a law professor specializing in nonprofit law at Northwestern University, expressed puzzlement at Musk's standing: "The idea that Elon Musk can sue because he was a donor or used to be on the board is pretty puzzling" [23]. Chan Loui, another nonprofit law expert, argued that the court has been analyzing the claim under the law of trusts, when "OpenAI is not a trust. OpenAI is a corporation. And so really they should be looking at … the law of charitable nonprofit organizations" [23].

This distinction matters because trust law and nonprofit corporate law impose different obligations on organizational leaders and offer different remedies to aggrieved parties. If the court applies nonprofit corporate law rather than trust law, Musk's path to recovery narrows considerably.

The case is also expected to set precedent for how nonprofit-to-for-profit conversions are governed in the AI sector. California and Delaware attorneys general approved OpenAI's conversion in October 2025 [19], but a court finding that the conversion breached charitable trust could cast doubt on similar transitions by other mission-driven AI organizations.

OpenAI's Steelman: Survival Demanded Restructuring

OpenAI's strongest argument is a practical one: the nonprofit model could not sustain the organization's mission.

After Musk's departure in 2018, OpenAI faced a choice between irrelevance and adaptation. Altman testified that the company was "left for dead" by Musk's withdrawal of financial support [9]. Competing with Google Brain, DeepMind (backed by Alphabet), and eventually well-capitalized startups like Anthropic required billions of dollars in compute — resources no pure nonprofit could secure.

The capped-profit model introduced in 2019, and the subsequent PBC conversion in 2025, were presented as the only viable path to actually building safe artificial general intelligence, rather than merely aspiring to it [9]. Microsoft's $13 billion investment — and its projected $92 billion return — demonstrated that commercial partnerships could fund the research while maintaining a public-benefit mandate [16].

OpenAI also pointed to the structural safeguards in the PBC: the OpenAI Foundation retains a 26% equity stake and governance oversight, and the PBC is legally obligated to consider its impact on society, not just shareholder returns [19].

What Comes Next

Closing arguments are scheduled for May 15. The advisory jury is expected to deliberate and return a recommendation as early as the week of May 18 [2]. Judge Gonzalez Rogers will then issue the final ruling.

Whatever the trial court decides, the losing side is expected to appeal to the Ninth Circuit, where the case would likely be heard in late 2026 or early 2027 [2]. The appeal could take the form of a narrow challenge to the application of trust law or a broader test of donor standing in nonprofit governance disputes.

The case has already reshaped the conversation about AI governance, forcing a public accounting of how a research nonprofit funded by charitable donations became one of the most valuable private companies in history — and whether the people who built it honored, or betrayed, the promises they made along the way.