Sam Altman gave testimony addressing accusations over a charity and control of OpenAI’s future, while the case’s larger aim remains disputed.
Sam Altman’s courtroom performance leaned heavily on credibility and detail, but the fight at the heart of the trial may be bigger than any single witness getting through a cross-examination.
After about two weeks of testimony from a range of witnesses accusing Altman of dishonesty. the jury finally heard directly from Altman himself.. When his lawyer. William Savitt. asked him how it felt to be accused of stealing a charity. Altman responded that the charity had been built through hard work and that he agreed it couldn’t be stolen.. In the same answer, he also alleged that Elon Musk tried to “kill” the effort twice.
Altman delivered his direct testimony in a manner he framed as that of a “nice kid from St.. Louis,” at times appearing bewildered by the spectacle of the accusations.. He seemed nervous early on, before settling into a more fluent presentation.. When he left the stand holding binders of evidence. his demeanor was described as almost schoolboy-like. and overall the testimony was portrayed as credible—at moments even as something the jury appeared to respond to.
What made the process difficult to follow. in the reporting surrounding the trial. was how familiar the narrative became to those watching who had already seen related public disputes.. The article points to other sworn statements from figures in the broader orbit—highlighting examples where testimony seemed contradicted by later events.. It also notes the difference between Altman’s situation and those other disputes. emphasizing that documents existed that. in the writer’s view. backed Altman’s account “at least. mostly.”
Altman’s testimony brought the court back to the central issue of how a future for-profit structure was supposed to be handled after OpenAI’s success in Dota 2.. He said discussions for a for-profit arm began in earnest after that win and described Musk’s position as requiring total control over the initial structure.. Altman testified that Musk only trusted himself to make decisions that would be correct. and that Altman personally felt uneasy about Musk’s insistence on control.
According to Altman, the discomfort wasn’t only about Musk’s level of involvement.. It was also rooted in the idea that OpenAI’s mission should prevent any one person from controlling AGI.. The testimony ties that principle to lessons Altman said he learned elsewhere. including at Y Combinator. where power struggles and control retention structures—like supervoting shares—are common.
Altman’s account also included a “succession” conversation with Musk when he raised questions about what would happen to control at OpenAI.. When asked about succession planning in the event of Musk’s death. Musk reportedly said he hadn’t thought about it much. but that control might pass to his children.. The reporting characterizes that answer as “hair-raising” and contrasts it with Altman’s stated belief that control should not concentrate around a single individual.
A 2017 email from Altman to Shivon Zilis was referenced to bolster the argument that Altman was worried about control even then.. In that email. Altman wrote that he was concerned about the idea of any one person controlling the world’s first AGI and connected that worry to the reason OpenAI was started.. He also said he was open to “creative structures. ” implying—based on the narrative—that Altman may have been willing to grant Musk control tied to specific milestones rather than unlimited authority.
Altman returned to that theme when asked what he believed Musk wanted.. He testified that he believed Musk wanted long-term control and suggested that such control might have been achieved if the structure Musk preferred had been accepted.. The coverage links this account to later video testimony in the trial from Sam Teller’s deposition. described as indicating that Musk would not invest in anything he didn’t control.
The broader pattern of Musk’s control fixation. as described in the reporting. was traced back to Musk’s history with being pushed out of his own company—specifically referencing PayPal.. In this telling. the insistence on control becomes less a one-off dispute and more an ongoing strategy that reappears in later business conflicts.
The trial also included evidence described as showing Musk trying to recruit Altman to Tesla.. Text messages were shown between Altman and Teller. with Teller telling Altman that Musk was committed to building up Tesla’s AI and hoping that Altman. Brockman. and Ilya Sutskever would join.. Altman. in testimony. characterized this as a “lightweight threat”—the idea that Musk would pursue that AI plan inside Tesla with or without their involvement.
At the same time. Altman testified that Tesla was primarily a car company. and he argued that allowing Tesla to acquire OpenAI would betray OpenAI’s mission.. In a separate set of messages introduced through Teller’s testimony. Teller wrote to Zilis shortly after midnight on February 4. 2018. stating that he didn’t want OpenAI to continue without Musk and that he would rather disable it by recruiting the leaders.
Money and timing also surfaced in the proceedings.. The reporting says that when Musk stopped quarterly donations. OpenAI was left operating with a “shoestring” budget and an “extremely short runway of cash.” It further states that OpenAI had other donors who were not the ones suing or joining Musk’s legal effort.. The article names Alameda Research as a firm connected to Sam Bankman-Fried. who is described as being in prison for fraud and money laundering. and notes that Alameda Research appeared in an exhibit but wasn’t called out in the courtroom.
Musk’s resignation from the board was portrayed as something that left uncertainty about his next move.. The article says some people wondered whether Musk might seek “vengeance” against OpenAI.. Altman’s view. however. was that Musk demotivated key researchers and caused “huge damage” over a prolonged period to the organization’s culture. leading to a picture—based on the report—that some employees may have felt relief when Musk stepped away.
The coverage also takes aim at the quality of Musk’s side of the legal presentation. describing evidence that Musk and others were reacting in ways that felt inconsistent with earlier claims.. It says that while Altman was setting up OpenAI’s for-profit arm. Musk was kept apprised of what was happening either directly or through Zilis or Teller. and that Musk did not object during the process.. The article notes that whatever Musk said publicly about Microsoft investments, there was evidence that privately he had been informed.
On cross-examination. Altman faced a barrage centered on accusations that he had been called a liar by a long list of people.. The reporting describes more than ten minutes of questioning from Steven Molo. citing figures including Sutskever. Mira Murati. Helen Toner. Tasha McCauley. Daniela and Dario Amodei—former OpenAI employees who founded Anthropic—as well as employees from Altman’s first startup Loopt and even references to a recent New Yorker article and a book titled The Optimist.
The reporting says Molo scored points by asking Altman about testimony in the trial and by stressing that Altman claimed he wasn’t paying close attention to it.. Altman, described as keeping his cool, appeared hurt and confused by the focus on whether he was lying.. According to the narrative, this was also the most successful part of the cross-examination, before the questioning declined in momentum.
The article further criticizes specific cross-examination tactics. describing an argument that tried to build significance from Altman being both a CEO and on the company’s board.. Altman countered—accurately in the reporting—that CEOs are almost always on the boards of the companies they run.. The writer characterizes the moment as exposing weaknesses in the questioning.
While the courtroom drama unfolded around credibility and governance. the report also highlights what it describes as a broader theme: the point of the trial isn’t framed as winning a case in the conventional sense.. The coverage discusses an argument related to fundraising for nonprofits. including a claim that if Stanford could raise $3 billion a year. OpenAI should have stayed a nonprofit.
But the reporting pushes back on that reasoning by comparing institutions.. It notes that Stanford’s fundraising and capital needs differ from those of OpenAI. and that Stanford is not competing with reputable for-profit companies in the same direct way.. It also argues that. setting those differences aside. $3 billion would not have covered scaling OpenAI to where it is now. especially if compute is treated as a main bottleneck.. That framing is used in the piece to suggest the defense’s nonprofit argument does not match the realities of building advanced models.
Yet the larger claim in the coverage is that the trial’s purpose is not to win; it’s to punish.. In that view. the legal action has already worked as a reputational weapon—reinforcing in the public mind an image of Altman as a liar and a “snake.” The report says Musk has pursued that goal in the public sphere. and that the legal process may be only one part of a longer campaign.
The article adds that. in the morning it was written. it read an exclusive report indicating that assorted Republican AGs and a House Oversight committee wanted to look into Altman’s investments.. It describes the trial as referenced throughout that later reporting. suggesting lawmakers and attorneys may be drawing from the court narrative even beyond the immediate dispute.
If there is a single uncertainty left after Altman’s testimony. it is the one the report itself frames as most troubling: he may have been convincing on the stand. and the suit’s outcome is still unknown. but it appears Musk’s effort to reshape public perception is far from finished.. For technology and AI watchers. the stakes go beyond courtroom outcomes. because the dispute sits at the intersection of AI governance. funding models. and who gets to define the rules for the next generation of powerful systems—questions that remain as live as ever.
Sam Altman trial OpenAI control AI governance charity allegations Elon Musk lawsuit courtroom testimony