ムスク対アルトマン、激突の幕開け

Elon Musk cofounded OpenAI, and then flounced off in a huff when he wasn’t anointed CEO, leaving Sam Altman as the last power-hungry man standing. Now, Musk is back with a lawsuit, and a trial is scheduled to start in Oakland, California, on April 27th. *Theoretically*, it’s a legal case about whether OpenAI defrauded Musk. But that’s not really what we’re all doing here. This is about mess.

Over the past couple of years, Musk’s legal theories for punishing OpenAI have run the gamut from breach of contract to unfair business practices to false advertising. Now, he and Altman will be getting called to the stand at a particularly delicate time. Musk’s xAI, now a part of SpaceX, has filed for an initial public offering. OpenAI is rumored to be considering an IPO itself. There are only billions of dollars at stake.

And so naturally, there’s a lot of internal tech gossip coming to light. Questions about Elon’s “rhino ket” use won’t make it into the trial… but it’s in news stories because it showed up in the docket. We got excerpts from OpenAI President Greg “What will take me to $1B?” Brockman’s diary. Mark Zuckerberg, not involved in the litigation at all, has had several embarrassing texts made public, like allegedly telling Musk that he had Meta teams “on alert to take down content doxxing or threatening the people on your [DOGE] team,” weeks after claiming he refuses to moderate based on government requests. Also, Musk thinks Jeff Bezos “is a bit of a tool.”

Lawsuits appear to be Musk’s preferred alternative to therapy

The fact that the case got to trial at all is a win for Musk, who seems to be trying to damage OpenAI’s reputation however he can — from lawsuits to general shit-talking to, apparently, a homophobic dossier on Sam Altman that’s getting passed around Silicon Valley by “Musk intermediaries.” *Musk v. Altman* “only ended up at trial because Elon Musk can pay his attorneys to argue a losing case,” said Sam Brunson, a professor of law at Loyola University of Chicago. “If I were doing this on contingency, I’d assume I wouldn’t be getting paid.”

Over the next few weeks, high-profile AI executives, such as Microsoft’s CEO Satya Nadella and CTO Kevin Scott, will likely testify. Former OpenAI executives, such as cofounder Ilya Sutskever and Mira Murati, former CTO, may be called. The former board members involved in Altman’s temporary 2023 ouster from his CEO role may also testify.

Musk v. Altman

A free, limited-run newsletter about the biggest moments from the Musk v. Altman trial.

Email (required)

Lawsuits appear to be Musk’s preferred alternative to therapy. Musk has sued perceived adversaries of his X social media platform, including a suit against a nonprofit that was dismissed as “baseless” and another against the firm that successfully made him follow through on his agreement to buy Twitter. Tesla and SpaceX are hotbeds of litigation. Let’s not go over the numerous family law matters that Musk is involved in due to his 14 known children.

Musk has actually filed four lawsuits against OpenAI. The first — for a breach of the founding agreement — was in state court, in 2024; Musk withdrew it immediately before a major hearing. We then got the current lawsuit, also filed in 2024, in which Musk alleged “Shakespearean” deceit. Several of its claims, including a laughable invocation of racketeering law, have been dismissed. Another suit, filed a year later, this time by xAI, accused Apple and OpenAI of engaging in anticompetitive behavior by making an agreement to exclusively put ChatGPT into iPhones. (The case is ongoing.) A fourth case accused OpenAI of poaching xAI employees and stealing trade secrets. It was dismissed.

“Since launching a competing artificial intelligence company, xAI, Musk has been trying to leverage the judicial system for an edge.”

In court starting next week, Musk will be making three main claims: that Altman and Brockman, *et al.*, breached OpenAI’s charitable trust; that they participated in unjust enrichment (at Musk’s expense); and that they committed fraud. His lawyer will tell a jury that he was duped into giving OpenAI money on terms that Altman and Brockman didn’t live up to. Among other things, he’s demanding that Altman and Brockman be removed from their company roles, that OpenAI be required to award a certain amount of money to its nonprofit, and that it cease operating in its current structure as a public benefit corporation.

OpenAI has countered that Musk failed to prove that Altman and Brockman ever made him a “cognizable promise” that could amount to unlawful activity, and that he lacks the standing for some claims, among other objections. It’s pointed out that Musk could have intervened in the company’s 2025 recapitalization in the time between OpenAI filing and its review by two attorneys general, and he did not. “This suit is the latest move in Elon Musk’s increasingly blusterous campaign to harass OpenAI for his own competitive advantage,” OpenAI wrote in one filing. “Since launching a competing artificial intelligence company, xAI, Musk has been trying to leverage the judicial system for an edge. The effort should fail.”

In court, OpenAI could argue that it engaged in self-help — such as starting its for-profit arm — because Musk left it in the lurch when he pulled promised funding from the nonprofit, says Peter Molk, a professor of law at the University of Florida. But that may not be enough to protect OpenAI. “My walking away doesn’t mean you can break any agreement we have,” Molk says. Musk may argue that OpenAI should have brought him to court and forced him to pay up. Of course, if OpenAI had done that, it likely would have gone bankrupt.

The details that come out in the trial about OpenAI “will absolutely change its reputation, if it’s still trying to claim it’s doing this in some high-minded, ‘we want to make AI safe for humanity’ way,” said Deven Desai, a professor of business law and ethics at the Georgia Institute of Technology. “The court documents and testimonies will make it harder and harder for OpenAI to keep claiming that’s what it’s about.”

”I think at this point OpenAI has the leverage to ask for a soft promise for new investors not to invest in competitors.”

Since this lawsuit was initially filed, OpenAI’s reputation has cratered. Besides the assorted lawsuits from people who say ChatGPT encouraged loved ones’ suicides, constant exec reshuffling, and the dominance of Anthropic’s enterprise product have significantly cooled enthusiasm for the company. And there’s always the question of whether OpenAI, one of the most expensive startups ever, will make a profit that meets investors’ expectations.

Already, we’ve found out that Sutskever and others were worried about the success of Stability AI, a then-competitor open-source lab. Sutskever also felt “betrayed” by Reid Hoffman, an early investor, founding his own AI lab; in response, Altman said, “I think at this point OpenAI has the leverage to ask for a soft promise for new investors not to invest in competitors.” Altman also didn’t tell the board he was personally running an OpenAI VC fund, according to a deposition of former board member Helen Toner.

Some of the damages Musk is calling for in his lawsuit — like the demands to unseat executives and change the company’s business structure — are likely unrealistic. State attorneys general from California and Delaware both blessed OpenAI’s restructuring. But Georgia Tech’s Desai says that even if the federal court doesn’t move to act on such requests, Musk might still get what he wants. The suit could do real damage, especially ahead of OpenAI’s impending IPO, and amid some shareholders reportedly questioning if Altman is the person to lead the company during that process — especially as allegations of his untrustworthiness and manipulative behavior resurface.

Musk’s strategy here is likely not just about angling to win in court, but also to “go after OpenAI as it currently exists,” says Desai. “It’s to create enough issues around how it has evolved to cause trouble and possibly get to the point where even if Musk doesn’t win, he’ll have made it look like it’s not worth keeping Mr. Altman in his position.”

Any scandalous information about OpenAI’s C-suite could derail the IPO

Musk has his own AI project with xAI and is nakedly trying to damage a competitor, says Molk. Much of the damage — bad PR, the cost of litigation, and distracting Altman and other executives as they begin to prepare for an initial public offering — occurs outside the purview of the courtroom, Molk says. “As long as there is some credibility [to Musk’s case], the motivation doesn’t matter,” he says.

The suit is part of a yearslong feud between Altman and Musk, ever since their personal and working relationships both soured in a public way. Musk’s xAI is aimed at the same government contracts and consumer chatbot users as OpenAI. He also folded xAI into SpaceX, as he had planned to do with OpenAI and Tesla.

OpenAI has responded to Musk by blogging. According to OpenAI, Musk himself wanted a for-profit arm of OpenAI — and further, 50 to 60 percent of its equity, and to be its CEO. He also proposed merging OpenAI with Tesla. OpenAI’s lawyers, playing dirtier, asked in depositions about Musk’s purported escapades at Burning Man, including whether he’d ingested “rhino ketamine.” This fits with its legal tactics elsewhere — including subpoenaing nonprofits and requesting a full list of who attended a deceased 16-year-old’s memorial services.

OpenAI is racing both SpaceX and Anthropic for an initial public offering. And there’s more investor pressure than ever to generate revenue, after companies like OpenAI and Anthropic raised billions of dollars without generating profit. Reportedly, even OpenAI’s CFO doesn’t believe the company is ready to go public in 2026 due to its sizable spending commitments. Other executives have nevertheless expressed a desire to beat Anthropic to the public market. Any scandalous information about OpenAI’s C-suite could derail the IPO.

Zilis: “‘Relationship’ is a relative term. But there have been romantic moments.”

Musk is also potentially vulnerable to mud-slinging. Musk’s SpaceX has filed confidentially for an IPO, which may happen as soon as June. Anything revealed in the courtroom about xAI or Musk personally could potentially affect investor interest or confidence, particularly since the IPO may follow so quickly on the heels of the trial. We already know from texts surfaced in discovery that Musk was actively recruiting from OpenAI.

There are other topics — such as Musk’s relationship with Shivon Zilis, the former board member of OpenAI and mother of several of his children — that are particularly sensitive, too. Zilis, who was effectively OpenAI’s “Elon whisperer,” has already questioned what the meaning of a “romantic relationship” is in her deposition. (Zilis: ‘Relationship’ is a relative term. But there have been romantic moments.”) And in a 2018 text message, Zilis asks Musk if she should stay “close and friendly” to OpenAI to “keep info flowing.” She also told Musk that after he hung up on a call, Sutskever was “visibly devastated” and Musk could probably recruit him if he wanted to.

Another key witness on OpenAI’s list is Jared Birchall, Musk’s fixer — who, among other things, runs Musk’s family office. Birchall is likely privy to a number of Musk’s secrets, and his testimony could be damaging.

On the other hand, “I’m not sure how much reputation Musk has left to lose,” says Brunson.

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

• Elizabeth Lopatto
• Hayden Field

-

-

-

-

-

-

-

-

-