Elon Musk's Testimony in the OpenAI Trial Is His Own Worst Enemy

Elon Musk built his lawsuit against OpenAI on a narrative of principled departure — a co-founder who left when the nonprofit lost its way. His own courtroom testimony may have just demolished that narrative. According to The Verge’s reporting from inside the trial, defense cross-examination strongly implied Musk cut his quarterly payments to OpenAI only after failing to secure majority control, not on any point of principle.

The Self-Defeating Witness

Musk handed defense attorney William Savitt an obvious target during direct testimony: the declarations “I don’t lose my temper” and “I don’t yell at people.” Within hours, the courtroom watched him contradict both. Judge Yvonne Gonzalez Rogers — who repeatedly had to prompt him toward yes-or-no answers — said afterward he was “at times difficult,” adding that getting through his testimony was itself a management challenge. The Verge reports jurors were visibly exchanging glances during testy exchanges, and Savitt consistently surfaced gaps between Musk’s morning testimony and his earlier deposition.

A Power Struggle Dressed as Principle

The most damaging portion of cross-examination targeted Musk’s departure rationale. According to The Verge, he had sought four board seats and 51 percent of shares on an initial seven-seat board — outright majority control from the start. When that arrangement fell through, his quarterly payments to OpenAI stopped. That sequence directly undermines the lawsuit’s central theory: that Musk left as an act of principled dissent against mission drift rather than a lost boardroom negotiation.

The Karpathy Contradiction

The Verge also details Musk’s 2017 recruitment of Andrej Karpathy — described in the reporting as OpenAI’s second-best engineer at the time — to Tesla during the same period he was withdrawing his funding. As a sitting OpenAI board member carrying fiduciary duties, Musk acknowledged on the stand that he made no effort to retain Karpathy. His stated rationale — “I think people should have a right to work where they want to work” — was undercut by his own contemporaneous message: “In my and Andrej’s opinion, Tesla is the only path that could even hope to hold a candle to Google.”

Why This Matters

Regardless of the jury’s eventual verdict, this trial is setting precedents for how courts interpret founding agreements at AI nonprofit labs. The more immediate lesson: Musk built his case on a principled-exit narrative, and his own words — on the stand and in contemporaneous messages — are proving to be the most effective rebuttal OpenAI has.