
Musk Threatened to Hit Brockman. Zilis Says He Tried to Poach Altman. Week 2 Was Chaos.
Week 2 of Musk v. OpenAI produced diary entries, physical threats, and testimony that Musk tried to recruit Altman for Tesla AI.
The second week of Musk v. OpenAI delivered the kind of testimony that screenwriters would reject as too dramatic. Greg Brockman told a jury he thought Elon Musk was going to physically assault him during a 2017 equity dispute. Shivon Zilis, the mother of four of Musk's children and a former OpenAI board member, revealed that Musk tried to recruit Sam Altman to lead a new AI lab at Tesla. And Musk himself, two days before trial, texted Brockman: "By the end of this week, you and Sam will be the most hated men in America."
Welcome to the most consequential tech trial in a generation. And the biggest witnesses have not even taken the stand yet.
The Haunted Mansion Meeting
Brockman's testimony, delivered over two days from a federal courtroom in Oakland, painted Musk as someone who was never committed to keeping OpenAI a nonprofit. According to Brockman, the tipping point came in the summer of 2017, when OpenAI's AI beat the world's best Dota 2 players. Musk hosted a celebration at what Brockman called his "Haunted Mansion" near San Francisco. The house was splattered with confetti and cups. Actress Amber Heard, Musk's girlfriend at the time, served whiskey.
"Time to make the next step for OpenAI. This is the triggering event," Musk wrote in an email, according to MIT Technology Review's reporting on Brockman's testimony. Weeks earlier, Musk had said that if OpenAI achieved a major public milestone, it would be "time to create a for-profit."
Over the next six weeks, intense negotiations followed. Musk wanted majority equity in the new for-profit entity, the right to choose a majority of the board, and to be its CEO. When Brockman and co-founder Ilya Sutskever proposed equal equity for all founders, Musk went silent, said "I decline," then "stormed around the table."
"I actually thought he was going to hit me," Brockman told the jury. Sutskever had brought a painting of a Tesla as a goodwill gesture. Musk grabbed it and walked out.
Zilis Takes the Stand
Shivon Zilis, a venture capitalist and neurotech executive who served on OpenAI's board and has four children with Musk, testified on Friday. Her most explosive revelation: Musk had tried to recruit Altman himself to lead a new AI lab at Tesla. The attempted poaching undercuts Musk's central narrative that he was betrayed by Altman. If Musk believed Altman was dishonest and uncommitted to the mission, why try to hire him?
Zilis's testimony builds on a pattern from the first two weeks. Brockman also testified that Musk used OpenAI employees to work on Tesla's Autopilot system for free, a claim that The Guardian's courtroom reporter described as one of the trial's most damaging moments for Musk's side.
The Settlement Text
Perhaps the most revealing moment came from texts sent just before the trial started. According to Brockman, Musk messaged him asking if he would be interested in settling. When Brockman proposed both sides simply drop their claims, Musk replied: "By the end of this week, you and Sam will be the most hated men in America. If you insist, so it will be."
That text, presented as a court exhibit, was notable for what it revealed about motivation. Musk is asking the court to remove Altman and Brockman from their roles, unwind OpenAI's restructuring into a public benefit corporation, and award up to $134 billion in damages from OpenAI and Microsoft. But the settlement offer suggests something simpler: leverage.
Week 3: Nadella, Sutskever, and the Real Fireworks
The trial now enters its third week, and the witness list is staggering. Microsoft CEO Satya Nadella is scheduled to testify Monday. Ilya Sutskever, OpenAI's former chief scientist and the co-founder who led the brief boardroom coup against Altman in 2023, is expected to follow. Both testimonies could reshape the trial's trajectory entirely.
Nadella's testimony matters because Microsoft invested $13 billion in OpenAI despite internal emails, revealed earlier in the trial, showing deep skepticism from his own engineering team. What did he know, and when? Sutskever's matters because he sits at the intersection of every faction: he co-founded OpenAI with Musk, led the safety concerns that drove the board revolt, and ultimately left to start his own AI safety company. His loyalties are genuinely unclear.
The stakes are not theoretical. If Musk prevails, OpenAI's restructuring could be unwound, threatening its race toward an IPO at a valuation approaching $1 trillion. Meanwhile, Musk's own xAI, now folded into SpaceX, is itself expected to go public as early as June at a target valuation of $1.75 trillion. The two richest men in AI both need a specific outcome from this trial, and the jury gets to decide which one.
The Guardian's Dara Kerr, reporting from inside the courtroom, compared it to a tech-industry remake of The Bonfire of the Vanities: "ambition, ego, greed and the spectrum of social class on full display." After two weeks, it is hard to argue with the comparison.
Reporting via MIT Technology Review, The Guardian, and Business Insider.