Steven Molo stood at the podium in a San Francisco courtroom and introduced a witness with the wrong name. The lawyer, leading Elon Musk’s case against OpenAI, called Greg Brockman — OpenAI’s president and an early Musk ally — “Greg Altman.” Later he mixed up dates on key evidence. The Verge, which covered the session, described the performance as “an incredible demolition derby.”
The moment captured the ragged end of one of the most closely watched trials in AI’s short history. On May 14, lawyers for Musk and OpenAI delivered closing arguments in the case, in which Musk claims the company abandoned the founding promise that made him an early donor and co-founder. The central question for the jury: whether OpenAI’s conversion from a nonprofit to a capped-profit company violated commitments its founders made in 2015.
Musk’s team argued betrayal. The company was founded as a nonprofit devoted to open, safe AI for humanity, they said; the shift to a for-profit structure with a profit cap, completed over several years, redirected the company’s loyalty to investors. OpenAI’s lawyers countered with survival. Without the ability to raise enormous sums — and to compensate researchers with equity — the company could not have competed for talent or compute, they argued, and the nonprofit mission would have died quietly alongside it.
The trial has pulled back the curtain on OpenAI’s early years in unusual detail. Jurors heard internal communications from the company’s founding period, including messages among Musk, Brockman, Altman and Ilya Sutskever, the former chief scientist. Microsoft’s investment terms were aired in open court, as were the disagreements among founders about control and direction. Musk, who left OpenAI’s board in 2018 after proposing to merge the company into Tesla, has said he watched from the outside as the mission drifted from its stated purpose.
The closing arguments were not without their own drama. Molo, a veteran litigator from New York, stumbled repeatedly on names and chronology, giving OpenAI’s lawyers room to argue that the case rested on a shaky factual foundation. Court observers said the mistakes undercut a narrative that had already struggled to find a crisp legal theory. OpenAI’s lead counsel, by contrast, walked through the company’s funding history methodically, arguing that every structural change was disclosed, approved by the nonprofit board and necessary to attract capital.
The case matters beyond the two men at the center of it. AI companies have modeled their governance on OpenAI’s capped-profit structure, which pairs a nonprofit board with a for-profit arm designed to attract venture capital. A jury finding that the architects of that structure misled early donors would ripple through boardrooms from San Francisco to London, lawyers said. A victory for OpenAI would hand the model a judicial seal of approval that other AI companies could cite when defending their own structures.
Both sides used their final hours at the podium to frame the stakes. Musk’s lawyers described OpenAI as a charity that became a corporation while promising its donors it would not. OpenAI’s lawyers described a startup that had to grow up or die, and noted that Musk himself later founded a competing company, xAI, that adopted the same for-profit form he now condemns.
The jury began deliberations after the arguments concluded, and the case now rests in its hands. Whatever it decides, the trial has already become a reference point for how AI companies structure themselves. Regulators in Washington and Brussels have watched the proceedings for evidence about concentration of power in AI, and the documents disclosed in court have been combed by antitrust officials, according to people familiar with the matter.
The trial also exposed how personal the dispute has become. Emails introduced in evidence show Musk and Altman once planning the company’s future together, trading ideas about what artificial general intelligence would look like and who should control it. Years later, they faced each other through lawyers in a courtroom, the founding friendship reduced to exhibits and objections.
For the jury, the question is narrow but consequential: did OpenAI break a promise? For the rest of the industry, the question is broader — what does a company owe the people who believed in its founding mission? Analysts said the answer will shape the governance documents of every AI startup founded in the next several years, as founders and investors try to write contracts that avoid the ambiguity now being litigated in public.
The episode fits a pattern that has defined the company’s approach through this cycle: move fast on consumer products, resolve legal questions later. Investors have mostly rewarded the strategy, though the open cases carry real financial risk. Legal costs, settlement payouts and management attention all flow to the same ledger, and each new filing extends the timeline before the underlying business question gets answered cleanly.
