Closing arguments wrapped this week in the Musk vs. Altman trial. A nine-person jury in a federal courthouse in Oakland, California, is now deliberating on whether Musk’s fellow cofounders defrauded him when they steered the company away from its nonprofit roots. The jury’s verdict is advisory. Whatever it returns, Judge Yvonne Gonzalez Rogers can set it aside entirely.
For founders globally, the stakes here are not abstract. An increasing number of startups building on OpenAI’s API are essentially dependent on the stability of OpenAI’s corporate structure. If Rogers rules that OpenAI’s conversion breached its founding terms, the resulting legal and governance disruption hits every developer and company built on top of it.
READ: Musk vs OpenAI: Week 2 Proved Control Was Always the Point
Why Sam Altman’s Truthfulness Became the Central Issue
Musk’s attorney Steve Molo spent considerable time in closing arguments questioning whether Sam Altman has been straight with the public. The sharpest exchange concerned testimony Altman gave to Congress, where he said he held no equity in OpenAI. He did hold a stake through Y Combinator, which he previously led.
Altman’s defence was essentially definitional: a passive interest in a VC fund is different from direct equity. He has acknowledged being conflict-averse, willing to tell people what they want to hear, and framed this as something he is working to change.
The question of Altman’s credibility extends beyond this courtroom. Journalists, policymakers, and consumers have been asking it for months. OpenAI is a privately held company. The only window into its intentions, outside of leaked documents, is what its leaders say publicly. When those statements don’t hold up under cross-examination, the credibility of everything else they’ve said starts to slide.
READ: Musk v. Altman Trial: Week One Exposed His Weakest Points
Musk’s Own Record Complicated His Case
Elon Musk’s legal team had a problem Molo acknowledged implicitly: their client’s own relationship with accuracy is well-documented. Musk’s testimony was combative on the stand. It also required him to correct the record on statements he had previously posted on X.
One post will likely haunt him. In 2020, Musk wrote that OpenAI was “essentially captured by Microsoft.” OpenAI’s lawyers pointed to that tweet in closing arguments as evidence that Musk understood the nonprofit’s relationship with Microsoft years before he filed his 2023 lawsuit.
Timing matters here. California’s statute of limitations is three years. If Rogers determines Musk knew about the alleged breach before August 2021, his case has no legal footing regardless of the merits. Musk is seeking $150 billion in remedies. Rogers is not obligated to award him a dollar. She will convene separate hearings on remedies if both she and the jury agree that OpenAI breached its charitable trust obligations.
READ: Meta Wants You to Trust Its Private AI Chat After Killing Encrypted DMs
What the Trial Exposed That the Industry Would Rather Not Discuss
Strip away the personal history, co-founders, a falling out, competing companies, and a structural question remains. Who holds these organisations accountable when their stated missions change?
OpenAI started as a nonprofit with an explicit commitment to benefit humanity. It is now a capped-profit entity with a multi-billion-dollar commercial business. The trial did not set out to examine whether that transition was good or bad for the world. But it forced both sides to explain, under oath, what they promised and what they delivered.
The trial produced no clean villain. Musk was not in court for the closing arguments. His attorney, Molo, apologised for the absence; OpenAI’s lead counsel, William Savitt, did not let the moment pass quietly.
Musk’s team called it “stealing a charity.” OpenAI’s team called it “sour grapes.” Everyone left the courtroom slightly worse than when they entered.
The institutions building the models that will run the next decade of technology are asking for public trust they have not demonstrated publicly. They are privately held. Their governance decisions happen behind closed doors. A federal judge in Oakland is currently the closest thing the industry has to an external check. That is not a comfortable position for anyone building on top of these platforms.
Don’t miss what’s next in AI. The GizPulse newsletter breaks down the stories that matter for founders and global tech watchers every week.
