The Elon Musk v. OpenAI Trial: The Scenario No Artificial Intelligence Could Have Written

The trial between Elon Musk and OpenAI, which opened on April 28 in a federal court in Oakland, California, has two dimensions.

It is certainly the climax of the personal conflict between Elon Musk and Sam Altman. But it also concerns a deeper question: Did OpenAI betray its founding mission as a nonprofit organization acting for the benefit of humanity? The problem, as we will see in this article, is that the clash of egos corrupts the underlying issue.

When OpenAI was created in December 2015, it presented itself as a nonprofit artificial intelligence research company whose goal was to advance digital intelligence “in the way that is most likely to benefit humanity as a whole,” without being constrained by the obligation to generate a financial return. Later, OpenAI’s charter would reaffirm that objective by stating that “our primary fiduciary duty is to humanity.”

It is on these founding principles that Elon Musk now relies in making his case. In his version of OpenAI’s history, Sam Altman, Greg Brockman¹, and he co-founded OpenAI around a promise: To develop cutting-edge artificial intelligence within a nonprofit framework that prioritized the public interest. In the federal complaint he filed in August 2024, Elon Musk explains that Sam Altman and Greg Brockman “wooed” and “deceived” him by presenting OpenAI as a nonprofit, open-source counterweight to the giants of Silicon Valley, first and foremost Google DeepMind.

Accordingly, OpenAI’s leaders obtained from him, on the basis of an altruistic mission, financial support (amounting to $38 million), as well as his help with strategy, partnerships, and recruiting. They then shifted the Company toward a structure incorporating a commercial logic. Elon Musk’s complaint also targets a number of OpenAI’s supporters, including Microsoft.

Sam Altman and Greg Brockman obviously dispute this account of events. They maintain that Elon Musk had long been aware of the discussions they had about the need to put in place a for-profit structure capable of raising the sums required for the development of high-level artificial intelligence. Elon Musk allegedly even took | part in those discussions before withdrawing from them when he realized he would not obtain a majority stake, absolute control, and the CEO position of the entity being contemplated. During the first days of the trial, Elon Musk in fact acknowledged that he had been aware of early discussions about converting to a for-profit structure, while saying that Sam Altman had reassured him that OpenAI would remain a nonprofit organization, something Altman had publicly and clearly stated.

Ultimately, what should strongly shape the arguments, from a purely factual standpoint, is Greg Brockman’s diary, whose contents were at the center of one trial session. In it, in November 2017, he wrote that he was “in favor of stealing the nonprofit foundation from [Elon Musk] and converting it into a certified B Corp² without him.” He also included this damning comment: “The narrative of [Musk] will rightly be that we were not honest with him, ultimately, about the fact that we always wanted to do for-profit, just without him.”

Moreover, Greg Brockman had to admit that, like Sam Altman, he had already invested in the chip startup Cerebras when OpenAI discussed a possible acquisition of the Company, and that he had never disclosed that investment to Elon Musk. That acquisition plan never came to fruition, but OpenAI did sign a partnership agreement with Cerebras, which significantly increased its valuation and, by extension, the value of Sam Altman’s and Greg Brockman’s investments. These private financial ties are difficult to reconcile with a governance focused exclusively on the public good.

Better or worse, Greg Brockman also had to acknowledge that he never fulfilled his initial promise to donate $100,000 to OpenAI’s nonprofit foundation at the time of its creation, but that he now holds a stake in the for-profit company worth close to $30 billion. That enrichment answers the question he asks himself in his diary about what he must do to reach a personal fortune of $1 billion.

More broadly, and down to the smallest details, Greg Brockman’s testimony lends credibility to Elon Musk’s accusations.

These substantive elements are compelling from a common-sense standpoint, which does not mean they are legally unassailable: One must therefore refrain from predicting the outcome of the trial. Caution is all the more necessary because the factual evidence underpinning the accusations against Sam Altman and Greg Brockman is tainted by the man presenting them. Elon Musk is probably the least suitable person to take up arms against his enemies: His deeply damaged image, because of his ethical excesses, his interest in weakening a competitor (the outcome of the trial could undermine OpenAI’s governance and disrupt its IPO), and his ego conflict with Sam Altman make him himself highly suspect.

He did not help his own case, moreover, by contacting Greg Brockman two days before the trial to explore a possible settlement. When Brockman suggested that both sides drop their respective claims, Musk allegedly replied that, by the end of the week, Brockman and Sam Altman would be “the two most hated men in America.”

One of the most famous aphorisms in contemporary American politics is Newt Gingrich’s line that “when your house is on fire, you don’t ask the firefighter whether he is faithful to his wife.” This professed willingness to overlook the moral virtues of political leaders also seems likely to shape the outcome of this trial, absent further revelations in the proceedings ahead, because the spotlight is at least as much on the accuser as on the accused, if not more so.

Still, on the merits, this trial once again highlights the boundless manipulative tendency of OpenAI’s leaders. But the fact that they may have betrayed OpenAI’s original mission does not necessarily mean that they betrayed Elon Musk³.

And that is precisely what the trial must establish.

—

¹ Greg Brockman is another co-founder of OpenAI and is now its president.

² This is a certification for for-profit companies that claim to have a positive impact on Society. It therefore does not correspond to a nonprofit structure.

³ The trial concerns an alleged injury to Elon Musk and the legal consequences of what he claims was promised to him and not delivered. It is not solely about the question of principle of OpenAI’s betrayal of its founding mission. But that issue is a central part of Elon Musk’s argument, because it serves to show what he believed he was supporting and how OpenAI’s subsequent trajectory would amount to deception.