OpenAI used Musk's own control demands against him in court
Elon Musk's lawsuit against OpenAI matters because it could still reach into who controls one of the most important AI companies. On day three of the trial in Oakland, California, OpenAI spent less time defending its origin story and more time trying to show the jury that Musk's complaint is really about power: he wanted more control, lost that fight, and is now calling the company illegitimate.
The most important change from Tuesday's testimony was not Musk snapping that OpenAI lawyer William Savitt was trying to trick him. It was Savitt using Musk's own past words to argue that Musk once backed a for-profit structure with a nonprofit beside it, then turned against OpenAI later, after the company grew without him. That gives OpenAI a cleaner defense than saying the company changed. It lets the company argue that Musk changed.
According to Reuters, Musk told the jury that Savitt's questions were "designed to trick" him during cross-examination in federal court. Reuters also reported that Savitt questioned Musk about a 2015 email in which Musk suggested OpenAI could be structured as a for-profit corporation with a parallel nonprofit. That matters because Musk's case turns on the claim that OpenAI betrayed its original charitable mission when it moved toward a profit-seeking structure.
OpenAI's lawyers pushed the same point from another direction. According to PBS NewsHour, citing the Associated Press, Musk testified that he initially wanted a majority stake in OpenAI and control of four of seven board seats. PBS also reported that Musk said he invested about $38 million in OpenAI between December 2015 and May 2017. For OpenAI, those details help recast him less as a betrayed co-founder and more as a founder who expected control and never got it.
That framing matters beyond courtroom theater because Musk is asking for extraordinary relief. Reuters reported that Musk is seeking $150 billion in damages from OpenAI and Microsoft, with any award going to OpenAI's charitable arm. The case also challenges the governance structure that OpenAI says now keeps its nonprofit parent in charge. On its official structure page, OpenAI says the OpenAI Foundation can appoint all members of OpenAI Group's board and replace directors at any time. It also says the foundation holds a 26 percent equity stake and a warrant for more shares if valuation milestones are met.
Tuesday's cross-examination also gave OpenAI a way to attack Musk's credibility on safety, which is one of the moral pillars of his case. According to CNBC's courtroom live coverage, Savitt argued that Musk does not know what OpenAI is currently doing on safety, and Musk said he did not know what OpenAI may be doing with respect to safety. CNBC also reported that Musk said Microsoft's $10 billion investment was the tipping point that convinced him OpenAI had violated its nonprofit mission. OpenAI's answer is simple: if Musk is judging the company's present mission, but admits he does not know what it is doing now, the jury has a reason to discount that judgment.
The skeptical case is still real. Embarrassing emails and sharp exchanges do not settle whether OpenAI violated charitable obligations or misled donors, employees, or the public. The Guardian reported that the trial continues to center on Musk's allegation that Sam Altman and OpenAI abandoned the nonprofit mission they originally sold. A rough cross-examination can weaken a witness. It does not automatically decide the law.
Still, OpenAI found a more dangerous argument for Musk than "you are wrong." It is arguing that he wanted the company, failed to get it, and is now asking a court to unwind a governance system that still matters for OpenAI's future. What to watch next is whether OpenAI can keep making that control story feel more concrete than Musk's mission story. If it can, the trial starts to look less like a charity case and more like a founder breakup with $150 billion and OpenAI's corporate structure hanging on the outcome.