Elon Musk and Sam Altman are on opposite sides of a federal civil trial in Oakland, California, over whether OpenAI’s leadership betrayed a charity-style mission when it embraced investor capital, commercial licensing, and a reorganized corporate stack. U.S. District Judge Yvonne Gonzalez Rogers is presiding; a nine-person jury heard opening arguments in late April 2026 after jury selection concluded, with schedules filed on the docket pointing toward finishing the liability segment before late May 2026. The case matters less for tabloid scorekeeping than for precedent: courts rarely get a clean look at whether a research nonprofit’s pivot toward capped profit can trigger charitable-trust scrutiny at this scale.
What Musk says happened
Musk helped launch OpenAI in 2015 and funded early years before departing the board in 2018. His complaint, first filed in 2024, portrays a founding bargain—open, safety-conscious research for humanity—that he argues was abandoned when the company tightened ties with Microsoft and raced to monetize large language models. Musk later founded xAI as a competitor, a fact defendants highlight when arguing motive. Musk’s team has asked the court to consider structural remedies, including leadership changes and unwinding parts of the recent recapitalization, not only money damages.
What is actually on the verdict form now
The docket reflects a heavy pruning exercise: dozens of early theories were dropped or trimmed so the jury focuses on two causes—unjust enrichment and breach of charitable trust—according to pretrial orders mirrored on the public recap archive. That narrowing is typical before a complex Silicon Valley trial: judges want jurors to decide discrete questions of duty and benefit rather than slog through overlapping tort labels.
How the trial is staged
Judge Gonzalez Rogers split the work into two phases. The first centers on liability—whether defendants wronged the nonprofit mission or Musk’s donor expectations—and the jury’s answers there are advisory; the judge retains final call on mixed questions of law and fact. A later remedies phase, if reached, would translate any liability finding into dollar disgorgement, injunctive relief, or corporate restructuring under her supervision. Pretrial scheduling orders on the docket also allocated substantial presentation clocks—about 22 hours for Musk’s side versus OpenAI defendants and about 5 hours for Microsoft—signaling how long each narrative was expected to run once witnesses began.
Money on the table—and where it would go
Court-adjacent summaries of January filings described Musk’s side seeking up to $134 billion framed as disgorgement of wrongful gains tied to the restructuring, with counsel signaling in open court that such sums would be directed toward OpenAI’s nonprofit foundation rather than Musk personally if awarded. Those headline numbers coexist awkwardly with valuation arithmetic: OpenAI has published eye-watering private-market capitalizations after 2025–2026 funding rounds, but a jury award in the hundreds of billions of dollars would still collide with corporate finance reality, enforceability, and cross-border investor rights. Treat any top-line demand as advocacy until a written judgment exists.
Microsoft’s separate headache
Microsoft remains a defendant on theories that its multibillion-dollar partnership helped cement a for-profit path. Executives including chief executive Satya Nadella were listed as potential witnesses in pretrial witness outlines circulating on the docket. Microsoft has denied wrongdoing in public statements mirrored in trade coverage. Even a modest adverse finding could reshape how strategic investors contract with mission-driven labs that still sit under nonprofit parents.
How OpenAI and Altman frame the defense
Company messaging has repeatedly labeled Musk’s suit baseless, casting it as sour grapes from a departed co-founder who missed the commercial upside. On substance, defendants argue corporate documents—not informal emails—define fiduciary duties, that early donations were used for permitted research, and that California oversight of charitable conversions was followed. If jurors credit that line, the charitable-trust claim fails regardless of how sharp Musk’s rhetoric reads on social platforms.
Evidence themes jurors are weighing
Trial testimony reportedly spanned internal strategy decks, capitalization tables, and board minutes about when to ship closed weights to enterprise customers. Plaintiffs emphasize moments where leaders allegedly reassured Musk about nonprofit discipline; defendants counter that venture-scale compute required capital structures Musk refused to fund. Secondary exhibits—personal journals, text threads—are the kind of material that can swing credibility even when they do not change black-letter trust law.
Why the verdict form is advisory
Advisory jury findings are uncommon in consumer trials but show up in complex commercial matters where judges want a temperature read before crafting remedies that might affect third-party investors. Gonzalez Rogers can accept or reject jury answers with written reasoning, which also sets up clearer appellate issues than a pure bench trial would.
Bottom line
Musk and Altman are colliding in Oakland over whether OpenAI’s move toward capped-profit AI commercialization breached charitable duties and unjustly enriched insiders and Microsoft. The jury is hearing a narrowed two-claim case under a two-phase plan, with advisory liability findings first and judge-led remedies later if liability sticks. Advocacy around hundreds of billions of dollars in disgorgement still has to clear the jury, the judge’s post-verdict rulings, and any appeals before cash or structural relief is real.