Musk-Altman Trial Surfaces Competing Claims Over OpenAI Governance and Tech Industry Conduct

In a federal courtroom in Oakland, California, a nine-person jury began deliberating on May 19, 2026, in what both sides agree is the most consequential nonprofit governance case in American tech history [1]. At stake: whether OpenAI's transformation from a charity into an $852 billion corporation represents a legitimate evolution — or, as Elon Musk's legal team contends, the "looting" of a charitable trust [2].

The first phase of Musk v. Altman has produced three weeks of testimony, hundreds of pages of internal communications, and a procession of witnesses that reads like a who's who of artificial intelligence. The jury's verdict will be advisory; Judge Yvonne Gonzalez Rogers will make the final determination on liability [3]. But the legal and financial implications extend far beyond the two billionaires whose names are on the case.

The Founding Promise and Its Erosion

OpenAI was incorporated as a nonprofit in December 2015 with a stated mission to develop artificial general intelligence "for the benefit of humanity" [4]. Musk, Altman, and Greg Brockman were among its co-founders. Musk contributed approximately $38 million in donations [1].

In 2019, OpenAI created a "capped-profit" subsidiary, arguing it needed to raise venture capital to compete in a capital-intensive field. Under this structure, investor and employee returns were capped at 100 times their investment, with any excess flowing back to the nonprofit [5]. The nonprofit retained full governance control.

By October 2025, that structure was gone. OpenAI completed a recapitalization that converted the for-profit arm into a Public Benefit Corporation (PBC) called OpenAI Group PBC, with the newly renamed OpenAI Foundation holding approximately 26% of the company's equity — worth roughly $130 billion [6]. The profit cap was eliminated. The California and Delaware Attorneys General signed off on the deal, with California AG Rob Bonta stating his office "secured concessions that ensure charitable assets are used for their intended purpose" [6].

Critics were unconvinced. A coalition of consumer advocacy groups and philanthropic organizations argued the restructuring was "full of holes" [7]. The San Francisco Foundation led a petition requesting the AG take stronger action to protect charitable assets [8]. And one detail went largely unnoticed at the time: OpenAI quietly deleted the word "safely" from its corporate mission statement [9].

What the Jury Must Decide

The legal question before the jury is narrow but loaded: did Altman and Brockman commit a breach of charitable trust and unjust enrichment by reshaping OpenAI into a for-profit entity [3]?

Musk's legal team argues that his $38 million in donations were not unconditional gifts but were encumbered by a specific mandate that the organization remain a nonprofit [2]. Under California charitable-trust doctrine, such donations create a fiduciary obligation to use assets exclusively for their stated philanthropic purpose.

OpenAI's defense is straightforward: Altman and Brockman never promised to keep OpenAI a nonprofit, and the restructured entity remains mission-driven [10]. Altman testified that he never made such a commitment to Musk and that the nonprofit structure was "left for dead" by the time Musk departed [11].

A threshold legal question may prove decisive before any merits are reached. Most charitable-trust scholars hold that standing to enforce donor intent rests with state attorneys general, not individual donors [12]. If Judge Gonzalez Rogers agrees, Musk's case could be dismissed on procedural grounds regardless of the jury's advisory finding.

The Credibility War

The trial's most damaging testimony came not from Musk or Altman but from former OpenAI insiders.

Ilya Sutskever, OpenAI's former chief scientist and one of the most respected figures in AI research, testified that he spent roughly a year compiling a 52-page document for the board detailing what he described as Altman's "consistent pattern of lying" [13]. Sutskever said Altman's behavior included "undermining and pitting executives against one another," conduct he argued was "not conducive to any grand goal," including the safe development of artificial general intelligence [13].

Former board members Helen Toner and Tasha McCauley reinforced this picture. Toner described Altman as someone with "a habit of putting words in other people's mouths to get people where he wanted them to go" [14]. McCauley testified about "a toxic culture of lying" [14]. Former CTO Mira Murati also testified that Altman had been dishonest with her [14].

Altman pushed back during his own testimony, denying that he acted selfishly or "stole the charity" [15]. His legal team sought to frame the testimony of former board members and executives as the grievances of individuals who lost a power struggle during the chaotic November 2023 board crisis, when Altman was briefly fired and then reinstated amid employee and investor pressure [11].

Follow the Money

The financial stakes revealed at trial put a concrete figure on what insiders stand to gain from OpenAI's transformation.

Sutskever testified that his personal stake in OpenAI was worth approximately $7 billion as of May 2026, up from roughly $5 billion in November 2025 [16]. Trial evidence also showed that approximately 600 OpenAI employees cashed out a combined $6.6 billion through secondary share sales [17].

Source: CNBC, TechCrunch

Data as of Mar 31, 2026CSV

OpenAI's valuation trajectory tells the larger story: from an initial nonprofit with no commercial value, to $28 billion in April 2023, to $157 billion in October 2024, to $852 billion as of its March 2026 funding round [18]. That growth was fueled by a series of massive capital raises, including a $40 billion SoftBank-led round in March 2025 and a $122 billion round in early 2026 [19][20].

Altman himself does not hold direct equity in OpenAI — a fact he and his defenders have repeatedly emphasized [21]. He drew a salary of $76,001 last year [21]. However, court filings revealed that Altman holds more than $2 billion in stakes in companies that have business relationships with OpenAI, including $1.7 billion in fusion company Helion Energy, $633 million in Stripe, and $258 million in anti-aging company Retro Biosciences [22]. A plan to grant Altman a 7% equity stake in OpenAI — which at current valuations would be worth nearly $60 billion — has been discussed but not finalized [23].

OpenAI's roughly 7,200 employees hold compensation tied to the for-profit structure through Profit Participation Units (PPUs), which give employees a share of profits rather than direct ownership [5]. A typical package includes a base salary of about $300,000 and annual PPU grants valued at $500,000, vesting over four years [5]. If a court blocks or restructures the for-profit conversion, the value of those PPUs could be thrown into uncertainty — making employees a stakeholder group with interests distinct from both Musk and Altman.

Source: Various sources

Data as of Feb 1, 2026CSV

The Boardroom and the Bedroom

Among the trial's most unusual revelations was testimony from Shivon Zilis, a former OpenAI board member and current Neuralink executive. Zilis testified that Musk offered her his sperm after noticing she had no children, describing it as a "purely platonic IVF arrangement" [24]. The arrangement resulted in the birth of twins in 2021, while Zilis was still serving on the OpenAI board [24].

Zilis told the court that she and Musk "had agreed on complete confidentiality" about the arrangement due to security concerns, and that she informed the OpenAI board only after Business Insider learned of the relationship and planned to publish a story [25]. Her testimony raised questions about undisclosed conflicts of interest at the board level during a period when critical governance decisions were being made.

Separately, internal communications introduced at trial revealed that Microsoft CEO Satya Nadella had privately expressed concern about his company's dependence on OpenAI, writing in an email: "I don't want to be IBM and OpenAI to be Microsoft" [26]. The exchange surfaced roughly three years after Microsoft wrote its first $1 billion check to OpenAI.

How OpenAI's Conversion Compares

OpenAI's structural evolution is not without precedent in the nonprofit technology world, but the comparisons reveal how thin the safeguards have become.

The Mozilla Foundation, which wholly owns the for-profit Mozilla Corporation, represents the closest analogue. Both structures feature a nonprofit parent controlling a commercial subsidiary. But Mozilla's foundation president, Mark Surman, has publicly questioned whether OpenAI's nonprofit control is meaningful in practice [27]. A key difference: Mozilla Foundation retains 100% ownership of its for-profit arm, while OpenAI Foundation holds only 26% [6][27].

OpenAI's PBC structure is legally required to advance its stated mission and consider the broader interests of all stakeholders [28]. The Foundation appoints all members of the PBC's board and can replace directors at any time [28]. But critics point to a structural problem: seven of the nonprofit's eight board members also sit on the for-profit's board, and the nonprofit hired its own employees only a month before the trial began [14]. Its current activities are limited to grant-making rather than AI research [14].

The question is whether a 26% equity stake and overlapping board membership amount to genuine mission governance or a legal fig leaf. Consumer advocacy groups have argued it is the latter [29].

Available Remedies and Legal Precedent

If the jury advises a finding of liability, the remedies phase — which began in parallel on May 18 — will determine consequences [3]. Musk's legal team has sought several forms of relief: as much as $134 billion in damages to be awarded to the nonprofit, the removal of Altman and Brockman from their positions, and an injunction requiring OpenAI to honor its "original charter commitments" with continued court jurisdiction [30][3].

Whether any of these remedies are practically enforceable is an open question. There is no settled legal precedent for a charitable-trust conversion at this scale [12]. Musk himself testified that a ruling against OpenAI would set a precedent against "looting every charity in America" [2]. OpenAI's legal team has argued that the California AG's sign-off on the restructuring effectively resolves the charitable-trust question [6].

Legal scholars remain divided. Some argue that the AG's approval, while significant, does not immunize OpenAI from private claims of breach, particularly if the court finds that the approval was based on incomplete or misleading representations [12]. Others contend that allowing individual donors to override the AG's determination would create unworkable precedent.

The standing question looms large. If Judge Gonzalez Rogers determines that Musk lacks standing to bring charitable-trust claims — a power typically reserved for attorneys general — the case could end before the jury's advisory verdict matters [31].

What This Case Is Really About

Strip away the personalities and the case reduces to a question that will shape the next decade of technology governance: can a nonprofit convert itself into one of the world's most valuable private companies while keeping its charitable commitments meaningful?

The California and Delaware AGs said yes, with conditions. Musk says no, though his motivations are clouded by his ownership of xAI, a direct OpenAI competitor [32]. Four former OpenAI insiders testified under oath that the man running the company has a pattern of dishonesty [14]. And thousands of employees hold billions of dollars in compensation that depends on the for-profit structure surviving.

The jury in Oakland will offer its advisory verdict. The judge will decide. But regardless of the outcome, Musk v. Altman has already made the internal record public. The emails, the text messages, the 52-page dossier on a CEO's alleged lies, the boardroom sperm offer — none of that goes back in the box. The precedent it sets for nonprofit-to-for-profit conversions in technology will be studied and cited for years, whether the ruling ultimately favors the man who gave $38 million or the man who turned it into $852 billion.