The Emails That Undercut Elon Musk's Case: Trial Evidence Showed He Wanted the Same For-Profit OpenAI He Later Called Stolen

On May 18, 2026, a federal jury in Oakland, California, took less than two hours to reject Elon Musk's $150 billion lawsuit against OpenAI and CEO Sam Altman [1]. The jury found that Musk had waited too long to sue — that he knew or should have known about OpenAI's shift toward a for-profit model years before filing in 2024 [2]. But the statute of limitations ruling, while dispositive, was almost beside the point. Over three weeks of testimony, the trial produced something more consequential than a verdict: hundreds of pages of private emails, text messages, and internal meeting notes that told a story fundamentally at odds with Musk's public claims [3].

Musk had framed the lawsuit as a betrayal narrative — that he co-founded and funded OpenAI as a nonprofit dedicated to safe artificial intelligence, and that Altman and his allies had "looted" it for personal enrichment [4]. The trial record tells a more complicated story, one in which Musk himself pushed aggressively for the very corporate transformation he would later call theft.

The 2017 Emails: Musk's Own For-Profit Push

The most damaging evidence against Musk's narrative came from his own correspondence. In mid-2017, as OpenAI struggled to compete for talent and compute against deep-pocketed rivals like Google DeepMind, Musk began pushing the organization to adopt a for-profit structure [5]. The specifics of what he wanted were entered into evidence during cross-examination.

In fall 2017, Musk demanded majority equity, absolute control, and the title of CEO of any for-profit entity OpenAI might create [5]. OpenAI's leadership rejected these terms. In internal communications, executives expressed concern that Musk "could become a dictator" and sent him an email with the subject line "honest thoughts" laying out their objections [5]. The concern, as documented in evidence, was that as OpenAI made progress toward artificial general intelligence (AGI), Musk would hold unilateral control — a concentration of power the founders considered antithetical to the mission [6].

When his bid for control failed, Musk escalated. In September 2017, he wrote to Altman and others: "Guys, I've had enough. This is the final straw. Either go do something on your own or continue with OpenAI as a nonprofit. I will no longer fund OpenAI until you have made a firm commitment to stay or I'm just being a fool who is essentially providing free funding for you to create a startup" [5].

By January 2018, Musk proposed an even more dramatic step: merging OpenAI into Tesla [6]. When that too was rejected, he resigned from OpenAI's board in February 2018 [6].

During his own testimony, Musk acknowledged he was not opposed to a for-profit entity being created, "as long as it served the nonprofit and had a structure that capped the potential profit of investors" [4]. This is almost exactly what OpenAI did in 2019 when it created its capped-profit subsidiary, under which employees and investors would receive a capped return — initially set at 100 times their investment — with any surplus reverting to the nonprofit parent [7].

The Money: $38 Million of a $1 Billion Pledge

The trial also clarified the financial record. When OpenAI was announced in December 2015, its backers — including Musk, Altman, Greg Brockman, Reid Hoffman, Peter Thiel, and others — collectively pledged $1 billion [8]. Musk's individual pledge was widely reported as $1 billion, though the exact terms were never formalized in a written agreement, as Musk himself acknowledged on the stand [9].

Source: Bloomberg, NPR Trial Coverage

Data as of May 1, 2026CSV

The actual contributions fell far short. According to trial testimony and Bloomberg's reporting, Musk contributed approximately $38 million between late 2015 and May 2017 [9]. This included five quarterly grants of $5 million during 2016–2017, roughly $12.7 million in rent payments for OpenAI's office space between 2016 and 2020, and the provision of four Tesla vehicles to key employees [4]. No further cash contributions came after Musk's departure from the board.

Musk pushed back on the characterization that he had fallen short, arguing that he contributed "among other things, his reputation" and that his combined contributions of money and other intangibles exceeded $100 million [4]. There was no written contract governing his donations [9].

The Capped-Profit Conversion: Was It Legal?

In 2019, OpenAI created its capped-profit subsidiary, OpenAI LP, structured so that the original nonprofit retained governance control over the for-profit arm [7]. The question at the center of both the trial and a separate California Attorney General investigation was whether this conversion — and OpenAI's later plan to restructure further into a fully for-profit company — violated the organization's founding charter.

Legal scholars have been divided. UCLA Law professors Rose Chan Loui and Jill Horwitz argued that the "extraordinary circumstances" required for a nonprofit to change its purposes do not apply to OpenAI [10]. OpenAI's argument — that it cannot attract sufficient capital under a nonprofit structure to develop AGI — relies on the legal concept of "impracticability," a standard that is ill-defined and rarely applied, and their argument that investors dislike nonprofit constraints is unlikely to persuade a court [10].

The California Attorney General's office conducted a year-and-a-half investigation into the restructuring before reaching a settlement with OpenAI in October 2025 [11]. Under the deal, the newly formed OpenAI Foundation would retain approximately 26 percent of OpenAI's valuation — roughly $130 billion — with the ability to appoint board members and a special committee empowered to intervene on AI safety grounds [11]. Critics called the arrangement inadequate, noting that seven of the nonprofit's eight board members also sit on the for-profit's board, and that the nonprofit hired employees only a month before the trial began [3].

OpenAI's Fundraising Trajectory

The scale of capital that flowed into OpenAI after the 2019 conversion underscores why the governance questions matter. Microsoft's initial $1 billion investment in 2019 was followed by a $10 billion commitment in January 2023, and by mid-2026, Microsoft's total spending on OpenAI — including investment commitments, infrastructure, and hosting costs — exceeded $100 billion, according to Microsoft executive Michael Wetter's trial testimony [12].

Source: Reuters, TechCrunch, PitchBook

Data as of May 1, 2026CSV

SoftBank led a $6.6 billion round in October 2024 and followed with a $40 billion commitment in March 2025 [8]. At these valuations, the question of who controls OpenAI and what fiduciary duties exist to its original charitable mission is not academic — it is a question about the governance of one of the most valuable private companies in the world.

The xAI Problem: Musk's Own Governance Record

Perhaps the sharpest challenge to Musk's credibility came not from OpenAI's lawyers but from the public record of his own AI venture. Musk founded xAI in March 2023, initially incorporating it as a Nevada public benefit corporation (PBC) — a legal designation that commits a company to delivering social and environmental benefits alongside financial returns [13].

But Nevada public records show that as of May 9, 2024, xAI's PBC status was gone [13]. The change was made without public announcement. More striking: as recently as May 2025, Musk's attorney Marc Toberoff described xAI in an amended complaint as "a public benefit corporation founded by Musk to help accelerate scientific research via AI" — a full year after the designation had been terminated [13].

xAI has never published the annual environmental and social impact reports expected of a PBC under Nevada law [13]. And in February 2026, SpaceX acquired xAI in an all-stock transaction, structuring it as a wholly owned subsidiary [14]. Musk serves as founder, CEO, and Board Chair of xAI, concentrating operational and governance authority in a single person [14] — precisely the arrangement OpenAI's founders rejected in 2017 when Musk demanded it for himself.

The contrast is stark. OpenAI, whatever its governance shortcomings, operates with a nonprofit parent, a board with multiple independent members, and a capped-profit structure subject to regulatory oversight. xAI operates as a conventional for-profit subsidiary of SpaceX with no public benefit obligations, no external governance constraints, and unilateral control by Musk [14].

Who Testified and What They Said

The trial featured testimony from an unusual roster of tech industry figures. Microsoft CEO Satya Nadella testified that Musk never contacted him with concerns about Microsoft's investments in OpenAI violating any commitments [12]. Nadella's testimony was significant because Musk's lawyers had argued that the Microsoft partnership fundamentally compromised OpenAI's independence.

Five witnesses called Altman a "liar" under oath: Musk himself, former OpenAI chief scientist Ilya Sutskever, former CTO Mira Murati, and former board members Helen Toner and Tasha McCauley [15]. Sutskever testified that he had spent months gathering evidence of what he described as Altman's "pattern of deception and poor management" — the same concerns that led to Altman's brief ouster in November 2023 [15]. But Sutskever also said he reversed course and supported Altman's reinstatement because he feared the company would otherwise be "destroyed" [15].

OpenAI co-founder Greg Brockman's personal journal was entered into evidence, much to his dismay, revealing his private reflections on his evolving role at the company [3]. Texts between Musk and Meta CEO Mark Zuckerberg, in which they discussed possibly trying to buy OpenAI together, were also disclosed [4].

The Steelman Case for Musk

The strongest version of Musk's argument is not that for-profit AI development is inherently wrong — his own ventures make that position untenable — but that specific promises were made to him as a donor about how OpenAI would operate, and those promises were broken. Under this reading, Musk contributed money, reputation, and early strategic direction to an organization that explicitly committed to being a nonprofit serving humanity, and the people who took over that organization converted it into a vehicle for their own enrichment.

There is some evidence for this view. The five witnesses who called Altman a liar under oath suggest genuine governance problems at OpenAI [15]. The concentration of board overlaps between the nonprofit and for-profit entities undermines claims of independent oversight [3]. And the sheer velocity of OpenAI's fundraising — from nonprofit to $300 billion valuation in seven years — raises legitimate questions about whether the nonprofit structure was ever more than a fundraising strategy.

But the trial record also makes clear that Musk was not a passive donor who watched his charitable contribution get misappropriated. He was an active participant who pushed for the commercial transformation, demanded personal control over it, and left when he didn't get his way [5][6]. The jury didn't reach the merits, but the evidence suggests Musk's objection was less about whether OpenAI went for-profit and more about who got to run it when it did.

Precedent for the AI Industry

The case arrives at a moment when governance structures across the AI industry are under scrutiny. Anthropic, founded by former OpenAI researchers, operates as a public benefit corporation with its own hybrid governance model [16]. Google's DeepMind, originally an independent lab, was absorbed into Google's corporate structure years ago.

The California Attorney General's settlement with OpenAI will likely serve as a template — or a cautionary tale — for future nonprofit-to-commercial transitions in the sector [11]. The Harvard Law Review has identified what it calls "amoral drift" in AI corporate governance, where safety-oriented structures erode under commercial pressure [16]. Whether the OpenAI precedent accelerates or slows that drift depends on whether regulators and courts treat the settlement's terms as a floor or a ceiling.

Musk has vowed to appeal to the Ninth Circuit [17]. But regardless of the appellate outcome, the trial record is now public. The emails are out. And they tell a story in which both sides of Musk v. Altman wanted largely the same thing — a for-profit OpenAI — and fought primarily over who would control it.

What Remains Unresolved

The jury's statute of limitations ruling means the core legal question — whether OpenAI's conversion violated its founding charter — was never decided on the merits [2]. Judge Yvonne Gonzalez Rogers immediately adopted the advisory jury's verdict [17], but an appellate reversal could send the case back for a substantive trial.

Meanwhile, the broader governance debate continues. OpenAI's restructuring into a full for-profit entity, with the nonprofit foundation retaining a 26 percent stake, is proceeding under the terms of the California AG settlement [11]. Whether that structure adequately protects the original charitable mission — or simply dresses up a commercial enterprise in nonprofit clothing — is a question the trial raised but did not answer.