OpenAI Trial Week 1: Judge Shuts Down Musk as 7 Stumbles Undermine Case

Elon Musk spent three days on the witness stand in federal court in Oakland this week, and the judge made one thing clear: AI extinction rhetoric has no place in this trial. When Musk's attorney Steven Molo tried to argue that "we all could die as a result of artificial intelligence," Judge Yvonne Gonzalez Rogers fired back, pointing out the contradiction that Musk himself is building AI. "It's ironic your client, despite these risks, is creating a company that is in the exact space," Rogers said, according to NBC News. "There are some people who do not want to put the future of humanity in Mr. Musk's hands ... But we're not going to get into that business."

The rebuke was a recurring theme. According to Ars Technica, within minutes of the third day starting, Rogers had to instruct Musk to stop being sarcastic and evasive. He had been refusing to answer questions with anything other than: "You just can't steal a charity."

² catalogued seven specific moments where Musk's testimony damaged his own lawsuit:

• Concessions over objections: OpenAI lawyer William Savitt got Musk to make several key admissions despite his own lawyer's objections
• xAI's safety record exposed: Musk lost a fight to keep his own AI company's safety practices off the table, undermining his claim to be an AI safety advocate
• Contradicted by documents: OpenAI's lawyer showed emails, deposition testimony, and social media posts that contradicted Musk's testimony
• "Jackasses" remark: Musk was confronted with calling OpenAI's safety team "jackasses," per ²
• Safety cards ignorance: He admitted not knowing what "safety cards" are, despite his own company issuing them, Ars Technica reported
• "I don't yell": Musk testified he never loses his temper, then immediately raised his voice at OpenAI's lawyer
• Trump ties admitted: His lawyers failed to keep his connections to Donald Trump off the record

Musk's lawsuit hinges on the claim that OpenAI's co‑founders "stole a charity" by transitioning the company from a nonprofit to a for‑profit structure. He testified that he was made a "fool" and gave $38 million in "free funding" to start a nonprofit that CEO Sam Altman always intended to turn into an $800 billion company, according to Ars Technica.

But OpenAI's defense team painted a different picture. They argued Musk walked away after the other co‑founders refused to give him control over a proposed for‑profit arm. Crucially, Musk agreed on the stand that at the time, he thought creating that for‑profit arm was necessary for the good of the nonprofit — which is exactly how OpenAI positions its own for‑profit shift today. Musk also confirmed he never fulfilled his original $1 billion commitment to the company, NBC News reported, contributing only $38 million instead.

Sam Altman didn't bother showing up for Musk's first day of testimony but appeared for the cross‑examination, remaining "largely stone‑faced" throughout, The New York Times reported. The dynamic between the two men — once collaborators, now rivals — has become the defining narrative of the case.

OpenAI's lawyer Savitt, who previously helped Twitter executives force Musk to buy the social media platform, proved effective at pushing Musk's buttons. His opening statement set the frame: "We're here because Mr. Musk didn't get his way at OpenAI. Because he's a competitor, Mr. Musk will do anything to attack OpenAI," per.²

The four‑week trial could set a precedent for how AI companies restructure from nonprofit to for‑profit — a question with direct implications for builders. OpenAI completed its corporate restructuring in October 2025, shifting from a capped‑profit model to a more traditional for‑profit structure overseen by a nonprofit foundation. That for‑profit arm raised $122 billion in its latest funding round, NBC News reported. OpenAI plans an IPO in the last quarter of 2026.

If Musk wins, OpenAI could be forced to remain a nonprofit — potentially disrupting API access, pricing, and product roadmaps that millions of developers rely on. If OpenAI wins, the for‑profit conversion model gets a legal stamp of approval, which could encourage other AI labs to follow similar paths.

Meanwhile, Musk's SpaceX recently acquired xAI to create a combined private entity worth more than $1.2 trillion, NBC News reported. The judge's willingness to allow evidence of Musk's Trump ties suggests the trial will also explore whether political influence played a role in the competitive dynamics between the two companies.

During his testimony, Musk was asked about a claim he made last summer that xAI would soon surpass every company except Google. His response was revealing: he ranked Anthropic as the top AI provider, followed by OpenAI, Google, and Chinese open‑source models. He characterized xAI as a much smaller company with just a few hundred employees, according to TechCrunch.

This ranking is notable coming from someone suing OpenAI for straying from its nonprofit mission. It also acknowledges what many in the builder community already know: the AI landscape is more competitive than ever, and xAI is playing catch‑up despite Musk's public bravado.

The trial is scheduled to run for four weeks, with Musk's testimony now concluded. The case has a long way to go, but Week 1 established a clear pattern: Musk's emotional arguments about stolen charity and AI extinction have struggled to survive contact with documentary evidence and pointed cross‑examination.

For builders watching this case, the key question remains whether the outcome will destabilize the AI ecosystem they build on. An OpenAI forced back to nonprofit status would face an entirely different set of incentives — one that might prioritize safety over product velocity, but could also limit the compute and talent investments that power the APIs and tools developers depend on daily.