The end of the world will have to wait.

Over the course of three days of testimony in an Oakland federal courthouse, Elon Musk repeatedly mentioned his desire to avoid creating “a Terminator (opens in new tab) situation” with AI, a reference to the robot uprising in the 1984 film.

“In the movie … it’s not a good situation,” Musk said on the stand Thursday.

Judge Yvonne Gonzalez Rogers appeared fed up with Musk’s pontificating, particularly on the topic of killer robots. 

“It’s not the point of the case,” Rogers said as she instructed Musk that he was no longer allowed to bring up existential AI risk and the possibility of human extinction on the stand. “They got it. That’s enough.” 

The ongoing trial, in which Musk accuses his OpenAI cofounders of wrenching away control of the organization and converting it from a nonprofit to a for-profit company for the purposes of their own enrichment, is his show. 

Though his affronted attitude made it easy to forget, Musk is the plaintiff, not the defendant, and his testimony was his opportunity to tell his side of the story. Musk sought to lay out a simple but sweeping story, in which he gradually realized he had been duped by OpenAI executives Sam Altman and Greg Brockman, whom he repeatedly accused of “stealing a charity.” 

During his testimony, Musk argued that the future of all charitable giving in the U.S. is at stake, as is perhaps the fate of humanity, which he said could be imperiled by the technology OpenAI has built without him. But the judge made it clear that she intends to hold Musk and his lawyers to a much narrower set of claims, as she repeatedly ordered that questions about the well-being of human civilization and Musk’s relationship with President Donald Trump are irrelevant to the case. 

The nine-person advisory jury will eventually issue a ruling in the case, but it’s Rogers who will determine whether OpenAI’s current leaders unjustly enriched themselves by taking funding from Microsoft and moving away from the open, charitable structure initially envisioned by the founders. 

Throughout cross-examination, lawyers for the defense have tried a number of tactics to make Musk seem untrustworthy and motivated by jealousy over OpenAI’s success. OpenAI’s lead counsel, William Savitt, tried several times to catch Musk contradicting his own statements from depositions. Savitt also tried to demonstrate that Musk has at times seemed amenable, even supportive, of transforming OpenAI into a for-profit company, in sharp contrast to the case he’s making in court now.

OpenAI’s lawyers even attempted to undermine Musk’s claim that one of his primary concerns is that the company and its commercial interests will control a technology that poses an existential risk to humanity. 

At one point, an OpenAI lawyer asked Musk whether it was true that he had once tweeted, “The future is going to be amazing with AI and robots enabling sustainable abundance for all.” 

“Yes,” Musk replied. 

“And you believe the future is going to be amazing with AI and robots enabling sustainable abundance for all?” the lawyer continued. 

“Well,” Musk hesitated, “I’ve also said there are many possible futures. Some futures are good, and some are not good.” 

During jury selection, several prospects disclosed a personal antipathy or political opposition to Musk — so many that the judge acknowledged the reality that “people don’t like him.” But on the stand, Musk’s natural charms were on display, as evidenced by rounds of laughter emanating from the courtroom and the crowded overflow space. Musk often can’t resist a cocksure comeback or wry aside. 

“I”m not a lawyer,” he acknowledged after quibbling with a question, then added, smiling, “Well, I did take Law 101.” The crowd laughed. 

Rogers, who has presided over several high-profile technology cases, appeared to have little patience for Musk’s antics — or anyone else’s. She repeatedly instructed Musk to simplify his answers and demanded that lawyers stop interrupting him, stop asking leading questions, and stop getting distracted by issues not relevant to the case. “Sit down,” she yelled at one person in the final minutes of court Thursday. 

When she introduced the case Monday morning, Rogers reminded the court that it boils down to two allegations: “One is a claim for breach of charitable trust, and the second is a claim for unjust enrichment.” 

Also relevant is whether Musk waited too long to file his suit, missing the statute of limitations. OpenAI’s lawyers have highlighted moments when Musk could have intervened in the restructuring of the organization or expressed his dismay sooner. They argued that Musk didn’t sue until after founding xAI, a competitor to OpenAI. 

For his part, Musk repeatedly claimed he sued as soon as he was confident of what Altman and Brockman were up to. “No good, it turns out,” he quipped on the stand. 

Always a showman, Musk didn’t miss a chance to promote his business. Presented with the statement that ChatGPT has a technological lead on Grok, Musk’s competing chatbot from xAI, he said, under oath, that Grok is catching up. 

With many days left of the trial, it’s too soon to say which way the case is leaning. Other main players — including Altman and Brockman, who have appeared in court most days so far — have yet to testify. Brockman’s testimony in particular could illuminate how the other OpenAI founders decided to make deals with Microsoft and restructure the company. 

The trial will resume Monday, where the judge will continue her noble quest to keep robot armies out of the conversation.