Musk v. Altman: The Trial Is Louder Than the Case

Note: This post was written by Claude Opus 4.7. The following is a synthesis of reporting from CNBC, NPR, CNN, the Washington Post, Fortune, and Wikipedia’s coverage of the lawsuit’s procedural history.

For three days at the end of April, Elon Musk took the witness stand in a federal courtroom in Oakland, California, in a case he filed against Sam Altman, Greg Brockman, OpenAI, and Microsoft. He is seeking approximately $130 billion in damages and a court order returning OpenAI to nonprofit form. The headlines have written themselves — “looting the nonprofit,” “I was a fool” — and the trial has produced the kind of copy that financial-news editors live for. Beneath the spectacle, though, the case has problems the spectacle cannot solve.

What is actually being decided

Two procedural facts most coverage skips. First: the same judge presiding now, U.S. District Judge Yvonne Gonzalez Rogers, already ruled against Musk’s strongest pre-trial argument. In March 2025 she denied his motion for a preliminary injunction to block OpenAI’s for-profit conversion, finding that “Musk has not demonstrated likelihood of success on the merits.” A preliminary-injunction ruling is the judge’s first hard look at whether the plaintiff’s case can win on the merits. Hers said no.

Second: the jury that began hearing testimony April 28 is advisory only. Their verdict will not impose remedies on its own. Gonzalez Rogers — the same judge who has already publicly doubted the case’s merits — will issue the operative ruling. Even a unanimous “yes, OpenAI breached” from the jury would land with the judge whose first move was to say Musk was unlikely to win. That is the legal posture under the courtroom theater.

Musk on the stand

Three days. The lines that made the news on Day 1 (April 28): “I came up with the idea, the name, recruited the key people, taught them everything I know, provided all the initial funding.” Day 2 (April 29): the leadership was “looting the nonprofit”; he was “a fool” for funding OpenAI; Microsoft’s $10 billion investment was the “tipping point” that convinced him OpenAI had abandoned its nonprofit mission. Day 3 (April 30), under cross-examination by OpenAI’s lead trial attorney, Savitt, Musk grew combative — at one point telling Savitt that her questions were “designed to trick me, essentially” — and Gonzalez Rogers had to instruct him to actually answer the questions he was being asked.

The most quotable testimony made for excellent television. The most quotable testimony was also, by the time it reached cross-examination, running into a paper trail.

What the paper trail says

Savitt presented 2018 emails between Altman and Musk in which Altman explicitly told Musk that OpenAI was planning to raise additional funding from Microsoft, including a term sheet that contemplated a corporate structure raising up to $10 billion. The “I was deceived” framing — central to the breach-of-charitable-trust theory — has to coexist with documentary evidence that Musk was told what was being planned, in writing, eight years ago. He may not have liked the plan. He may not have been involved in the decisions that followed his 2018 departure from the OpenAI board. Neither is the same as having been deceived.

The damages math is also striking. Musk’s contributions to OpenAI from 2016 through 2020 totaled approximately $44 million. The $130 billion he now seeks is roughly three thousand times his actual contribution. The pleading directs the recovery to OpenAI’s nonprofit foundation rather than to Musk personally, which is procedurally clean — but Musk also runs xAI, which competes directly with OpenAI, and a $130 billion damages award plus the removal of Altman and Brockman from the board would be a competitive windfall for the plaintiff regardless of where the proceeds nominally land.

What it means for IT leaders

Two of the three AI vendors that an enterprise IT director realistically considers in 2026 — OpenAI and Anthropic — have active corporate-governance uncertainty. Anthropic is weighing a $900 billion private round while preparing for an IPO into a market valuing its secondaries at $1 trillion. OpenAI is in trial over whether its 2024 conversion to a public-benefit corporation was lawful, with its founding nonprofit retaining only 26% of the resulting OpenAI Group PBC, Microsoft holding 27%, and a federal judge presiding who has publicly doubted the case’s merits but who will nonetheless issue the operative ruling.

The litigation overhang is unlikely to result in OpenAI being unwound. It is also not nothing. Procurement decisions whose contracts run through 2027 are decisions about which vendor’s corporate form, board composition, and revenue-recognition rules might look meaningfully different in eighteen months. The reasonable answer is to keep the existing vendor relationships running and not architect for any specific outcome — but to read the structural-risk disclosures in your AI services contracts, because those tend to grow longer in the year after a trial like this one.

Bottom line

Trials produce great quotes. Cases produce rulings. The Musk v. Altman trial has produced great quotes. The case has produced one ruling so far — a preliminary-injunction denial in which the same judge who will rule on the trial’s outcome found that Musk had not demonstrated likelihood of success. The 2018 emails complicate the central deception claim. The advisory-jury structure means even a Musk-friendly verdict ends up back in front of the judge who already said no. The most likely final ruling is one that aligns with the preliminary ruling — and reads as a surprise only to readers who tracked the loudest moments of the trial and not the quietest.

What is actually being decided

Musk on the stand

What the paper trail says

What it means for IT leaders

Bottom line

Sources

Security Scorecard