Monday marked the first official day of the trial of Elon Musk’s lawsuit against Sam Altman and the artificial intelligence developer OpenAI, the company responsible for the creation of ChatGPT.
Musk claims that even though OpenAI was formed as a nonprofit corporation with a mission of benefiting humanity and not profits, under Altman’s leadership it has prioritized profits in violation of its charitable charter.
The federal judge trying the high-profile case in the federal courthouse in Oakland set the day aside for jury selection.
“There are a lot of people out there who don’t like your client,” U.S. District Judge Yvonne Gonzalez Rogers advised counsel for Elon Musk, in a courtroom exchange before jury selection began.
“And there are plenty who don’t like Mr. Altman,” she added.
The judge said she was looking for a jury that would carefully review the evidence and weigh it without being unfairly influenced by the public personalities of the two protagonists.
The jury will have an important but limited role. The case is being tried in two phases. The first phase will determine whether the defendants are liable to Musk, who claims that changing OpenAI from a nonprofit to a for-profit violated the organization’s original charter. If liability is proven, the second phase of trial will determine which damages and/or remedies are appropriate — or rather, what consequences the defendants will face.
The jury will hear the liability phase but not damages. Moreover, the jury is an “advisory jury” that will make their determination for the benefit of the judge. Gonzalez Rogers is not bound to follow the jury’s findings, though she has said in conferences with the parties’ lawyers that she likely will.
The judge led the process and questioned potential jurors.
Querying the jurors
Before arriving at the courtroom, potential jurors had filled out questionnaires designed to surface potential issues — financial, medical, conflicts or bias — that might affect their ability to serve.
A number of prospective jurors indicated that they had strong feelings about Musk, particularly his political activities, though most believed they could decide the case without being influenced by their opinions. One said, “The facts are the facts.”
Others expressed their views about artificial intelligence. Some were negative — one expressed concern about job losses — but a good number were using AI for personal matters and/or at work and generally found it a useful tool.
The judge gave the lawyers a chance to ask questions directly to the jurors.
Steven Molo, lead counsel for Musk, probed several jurors who indicated that they had a negative opinion of Musk, testing whether they would be able to put those aside. He used the opportunity to point out to the potential jurors that Musk’s political activity had nothing to do with the case.
One juror with a negative view of Musk also had strong views on what AI might mean in the future as governments become deeply involved. Molo wanted to know if the juror would be open to hearing that Musk shared a similar concern about AI’s dangers.
William Savitt, OpenAI’s top trial lawyer, focused on jurors who expressed admiration for Musk’s products. He solicitously thanked every juror that he questioned for their time. He asked one whether he considered himself an “Elon fan.” All said that they would not let those feelings influence their decision if selected.
Gonzalez Rogers is a highly experienced trial judge, after serving as a state court judge and before that a litigation partner at a well-respected San Francisco law firm. She is known by lawyers to be rigorous, highly prepared and insistent that she — not the lawyers — run her courtroom, a reputation shaped in part by her handling of cases like Epic Games v. Apple. In her questioning of potential jurors, she appeared friendly, occasionally making jokes and listening carefully to their responses to her questions.
After the questioning, the judge dismissed potential jurors she thought were not able to serve. She gave the lawyers an opportunity to challenge jurors on grounds that bias or personal circumstances made them unqualified to serve. Molo made three such challenges, and all were denied by the judge, who thought the jury would be able to put aside any preconceptions and resolve the matter impartially.
She let each side strike five witnesses “peremptorily,” that is by choice. Molo used all five; Savitt four.
When all was done, Gonzalez Rogers swore in a nine-person jury.
Protesters gather
Meanwhile, outside the federal courthouse, a small handful of protesters gathered — though they didn’t outnumber the press.
Many of their signs expressed fear of artificial intelligence. One sign depicted a cartoon Altman, saying: “Could [AI] destroy mankind?” Others bore phrases such as: “But rich people would save a bit of money, and that’s worth it” or “Creep State is Watching.”
A large, thin balloon of Elon Musk giving a Nazi salute also dominated the scene.
Latest case developments
Gonzalez Rogers issued two orders of note on Friday. First, she approved Musk’s decision to drop — irrevocably — two of the five counts that were set for trial. Second, she denied an emergency motion to intervene — that is, to join the case as a party — that was filed on March 31.
Musk’s lawyer said the decision to drop the fraud and constructive fraud counts was made to streamline the proceedings, though in the chess game of legal strategy there are other implications.
The decision whittles the case down to Musk’s claims for breach of charitable trust and for unjust enrichment against Altman, OpenAI and Gregory Brockman (with Musk and Altman, a co-founder of OpenAI) and against Microsoft for aiding and abetting the breach. Simply put, Musk is arguing that OpenAI’s leaders misused a charitable organization for private gain and Microsoft helped them do it.
The charitable trust claim is based on the core idea that OpenAI was chartered as a nonprofit corporation with the mission to develop artificial general intelligence (AGI) for the benefit of humanity, not for private profit. Musk contends that under Altman’s leadership, the company has been run in violation of the charter.
The claim for unjust enrichment is founded on the idea that the defendants have received money or property that they were not entitled to, and they should not be allowed to retain it.
As he put it, “despite solemn promises that OpenAI would operate as a nonprofit for the benefit of humanity” — and after accumulating billions of dollars as a charity — the defendants “converted those assets into a wealth machine for themselves, Microsoft, and Silicon Valley insiders.”
Paring the case down to the breach of charitable trust and unjust enrichment claims have potential advantages for Musk.
One important issue in the case is whether Musk brought his claim in a timely way in light of the relevant statutes of limitations. Breach of charitable trust has a four-year statute of limitations, longer than the dismissed counts. That may allow Musk to avoid spending time on issues that arose outside of the shorter limitations period, but within the four-year period.
Moreover, the dismissed fraud claims focus on what was said to Musk and whether he was misled by it. OpenAI is likely to press hard on the point that Musk’s sophistication and resources make him an unlikely person to be duped. Breach of charitable trust, however, is more focused on the defendants’ conduct and its fidelity to the trust’s requirements. That focus is more consistent with Musk’s narrative that he is pursuing the suit for the public interest.
The trial is scheduled to continue on Tuesday with opening statements and, depending on timing, testimony of live witnesses.
This story of part of a Bay City News special project following Elon Musk’s lawsuit against Sam Altman and OpenAI. The series explores the case’s ramifications for AI technology and the future of all humankind!