OAKLAND, Calif. — Elon Musk sat in the witness chair. His voice carried the familiar mix of certainty and urgency. Over two days he laid out a simple accusation. OpenAI had betrayed its roots.
“This lawsuit is very simple,” he said. “It is not OK to steal a charity.” The words landed hard in the federal courtroom. If allowed to stand, he warned, the move would “give license to looting every charity in America.”
The trial, now in its second week, pits the world’s richest man against one of his former collaborators. Musk co-founded OpenAI in 2015 as a nonprofit dedicated to safe artificial intelligence development. He invested roughly $38 million. Then he watched, he claims, as Sam Altman and Greg Brockman steered the organization toward a for-profit model capped by a lucrative Microsoft partnership.
But the testimony that followed Musk’s own time on the stand struck perhaps the sharper blows. Former insiders described a pattern of deception at the highest levels. They painted Altman as a leader who said one thing to one executive and the opposite to another. Chaos became routine. Trust evaporated.
Mira Murati, OpenAI’s former chief technology officer, delivered her account by video on May 6. She had briefly served as interim CEO after the board ousted Altman in late 2023. “My concern was about Sam saying one thing to one person and completely the opposite to another person,” Murati testified, according to a Reuters report. She accused him of sowing distrust, fostering infighting and creating “complete and utter chaos” that nearly tore the company apart.
Murati warned the organization stood at “catastrophic risk of falling apart.” Yet she had pushed for Altman’s return at the time. The alternative, she feared, looked worse.
Her statements echoed earlier depositions. Former board member Tasha McCauley described a “culture of lying and culture of deceit.” She pointed to repeated “crisis events” every few months. One involved the launch of GPT-4-Turbo. Altman allegedly told the board that OpenAI’s legal team had cleared it without need for safety review. That claim proved false.
Rosie Campbell, a former safety researcher at OpenAI from 2021 to 2024, watched the safety focus disappear. The company once maintained two dedicated long-term safety teams. One worked on aligning advanced systems with human values. The other prepared for superintelligent AI. Both teams were disbanded. About half her colleagues left rather than accept reassignment to product roles.
Campbell signed the 2023 letter urging Altman’s reinstatement. She worried the company would collapse without him and that talent would flow to Microsoft, which she viewed as even less committed to caution. The testimony, drawn from the second week of proceedings, comes from a detailed account in Business Insider.
Columbia Law School professor David Schizer, testifying as a nonprofit governance expert for Musk, drove the point home. “The board and CEO need to be partnering, working together, to make sure the mission is being followed,” he said. “If the CEO is withholding that information, it’s a big problem.” Products launched without board knowledge. Microsoft tested versions of models without safety checks. The pattern, in his view, violated the charitable purpose.
Musk’s own testimony mixed personal history with stark warnings. He claimed credit for naming the organization, recruiting key talent and raising early funds. The original vision countered Google. It rejected profit motives. Technology would be shared openly.
A 2015 conversation with Google co-founder Larry Page shaped his thinking. Page reportedly called Musk a “specieist” for prioritizing humanity over digital life forms. That exchange, Musk said, helped spark OpenAI.
He acknowledged early talks about a for-profit arm. “A for-profit was fine,” he testified, “as long as the tail did not wag the dog.” But equity demands from other founders grew excessive. The process annoyed him. He left the board in 2018.
Concerns about artificial intelligence run deep for Musk. “I have extreme concerns about A.I.,” he told the court. “It could kill us all.” He drew a contrast between dystopian outcomes and hopeful ones. “We don’t want to have a ‘Terminator’ outcome. We want to be in a Gene Roddenberry outcome, like ‘Star Trek.'”
The judge stepped in at one point. When Musk veered into extinction risks and an “AI army of robots,” Judge Yvonne Gonzalez Rogers cut him off. “We are not going to talk much about extinction in this case,” she said after sending the jury out. “They got it, that’s enough.” The moment, reported by The Guardian, underscored the court’s effort to keep the trial focused on contract and fiduciary questions rather than science fiction scenarios.
Cross-examination grew testy. OpenAI’s lead counsel William Savitt pressed Musk on details of the company’s structure and his competing xAI venture. Musk bristled. “Your questions are not simple,” he shot back. “They’re designed to trick me.” He repeated his core claim. “You can’t just steal a charity.” The judge struck the phrase from the record. The jury had heard it enough.
OpenAI’s defense painted a different picture. Musk simply “didn’t get his way,” they argued. He stayed silent on the 2019 Microsoft investment. Objections surfaced only after ChatGPT’s explosive success in 2022. By then Musk had launched xAI as a rival.
Emails shown in court revealed Musk once tried to recruit Altman to lead a Tesla AI effort and even offered him a seat on Tesla’s board. The overture, presented during testimony from Shivon Zilis, a former OpenAI adviser and mother to several of Musk’s children, suggested Musk once saw value in closer ties. OpenAI lawyers called it an attempt to absorb the lab into his own empire.
Jared Birchall, who manages Musk’s family office and serves as Neuralink CEO, followed Musk on the stand. He confirmed overseeing the $38 million in donations. All decisions came directly from Musk.
The trial has pulled back the curtain on Silicon Valley power dynamics. Private emails, text messages and diary entries have surfaced. They show the messy birth of a company now valued in the tens of billions. Greg Brockman, OpenAI’s president, took notes calmly while Musk testified. Altman entered and exited through a private door. Protesters gathered outside. Security remained tight after a recent firebomb threat at Altman’s home.
Recent coverage captures the shifting momentum. A May 6 Reuters analysis described the proceedings as a “soap opera” with world-altering stakes. Both men claim to protect humanity’s future. Each questions whether the other can be trusted with technology that may surpass human intelligence.
Former board member Tasha McCauley referenced a lengthy email from Ilya Sutskever cataloging dozens of chaotic incidents tied to Altman’s behavior. The document has not been fully public. Its existence alone fuels questions about oversight.
Campbell offered one telling detail. She viewed OpenAI’s safety culture as stronger than xAI’s, even while criticizing the company’s drift. The admission undercut any notion that Musk’s side holds a monopoly on caution.
Legal experts following the case say the nonprofit claims face steep hurdles. Courts rarely unwind corporate structures years after changes occur. Musk seeks $150 billion in damages to restore the charitable arm. He wants Altman and Brockman removed from control.
Yet the testimony has already damaged reputations. Altman’s image as a steady visionary has taken hits. Multiple insiders described systematic dishonesty. The safety team dissolution suggests commercial pressure won out over caution.
Musk, for his part, appeared rattled at times. He grew defensive under rapid questioning. His doomsday rhetoric, though curtailed, reminded observers why he left OpenAI in the first place. The man who runs Tesla, SpaceX and xAI still sees artificial intelligence as an existential matter.
The jury will sort through competing narratives. One side says Musk acts from sour grapes and thwarted ambition. The other insists a solemn promise was broken. A charity was taken over for personal gain.
Whatever the verdict, the trial has accomplished something rare. It forced the industry’s most powerful figures to defend their choices under oath. It exposed how quickly founding ideals can bend under success and capital. And it showed that even among those racing toward the future, old grievances refuse to stay buried.
Altman has yet to testify. His turn on the stand could reshape the story again. The courtroom remains packed. The lines outside have shortened as the trial stretches into its later days. But inside, the stakes feel as high as ever.
WebProNews is an iEntry Publication