The explosive legal battle surrounding OpenAI has taken a dramatic turn. In one of the most significant depositions of the ongoing Musk v. Altman trial, former OpenAI Chief Technology Officer Mira Murati testified under oath that CEO Sam Altman misled her about AI safety practices — accusations that are now part of the formal court record.
Who Is Mira Murati and Why Does Her Testimony Matter?
Mira Murati served as OpenAI’s CTO for years, making her one of the most powerful figures in the company’s history. She was briefly appointed interim CEO in November 2023 when the board temporarily removed Sam Altman, before he was swiftly reinstated. Her insider knowledge of OpenAI’s inner workings — and her willingness to speak on the record under oath — makes her deposition in the Musk v. Altman trial extraordinarily significant.
The allegations she raised were not entirely new. They had previously surfaced in a detailed New Yorker profile of Altman published earlier this year. But there is a substantial legal and reputational difference between a magazine account and sworn court testimony. Murati’s deposition transforms those claims into official, on-the-record evidence.
What Did Mira Murati Testify?
In a video deposition shown to the court, Murati stated that Altman falsely told her that OpenAI’s legal department had determined a new AI model did not need to go through the company’s deployment safety board. The model in question is widely believed to have been GPT-4 Turbo.
When an attorney directly asked her whether Altman was telling the truth when he made that statement, Murati answered simply: “No.”
She went on to explain that she had cross-checked Altman’s claim with OpenAI’s then-legal chief Jason Kwon — and found that what Kwon said and what Altman said were not the same thing. She described this as a “misalignment” between the two men. Crucially, Murati took matters into her own hands and ensured the model did go through the deployment safety board, overriding what Altman had told her.
“Creating Chaos”: Murati’s Broader Critique of Altman’s Leadership
Murati’s testimony extended well beyond the single safety dispute. She described Altman as “creating chaos” and said her central concern was that he was telling one thing to one person and the complete opposite to another.
She affirmed under direct questioning that Altman pitted executives against one another and actively undermined her ability to perform her role as CTO. Her criticism, she was careful to clarify, was entirely management-related — she described her role at the time as an “incredibly hard job” inside a complex and dysfunctional organization.
The picture she painted of OpenAI during that period was alarming. She said the company was “at catastrophic risk of falling apart” and that she was genuinely concerned it could completely implode.
Murati also testified that after Altman was reinstated as CEO following his brief November 2023 ousting, the problematic behaviors did not stop. She described continued delays on important decisions, inconsistent messaging to different colleagues, and a work environment that remained “very difficult and chaotic.”
The Musk v. Altman Trial: What’s at Stake
Murati’s deposition came during the second week of testimony in the Musk v. Altman federal trial, held in Oakland, California. Elon Musk filed the lawsuit in 2024, alleging that OpenAI improperly converted from a nonprofit into a for-profit company, abandoned its founding charitable mission, and should be forced to revert to its original structure.
In week one of the trial, Musk took the stand and argued that he had been deceived into providing early funding to OpenAI under the belief that it was a nonprofit dedicated to developing AI for the benefit of humanity — not for the financial enrichment of its executives. He described himself as a “fool” who provided tens of millions of dollars in essentially free funding that helped build what became one of the most valuable companies in the world.
If Musk succeeds, the outcome could fundamentally reshape OpenAI’s corporate structure, disrupt its path toward a potential IPO, and set legal precedents for how AI companies are governed going forward.
Other notable testimony has also emerged during the proceedings. Former board member Shivon Zilis testified that the board had voiced extreme concern about the release of ChatGPT without proper board communication — another signal that internal governance at OpenAI was under serious strain even before the 2023 crisis.
The 2023 Firing — and What Murati Did Next
The November 2023 boardroom upheaval at OpenAI remains one of the most chaotic episodes in Silicon Valley history. After Altman was removed by the board, Murati served as interim CEO for a handful of days, before Emmett Shear briefly stepped in, and then Altman was reinstated. Murati publicly supported his return at the time — a fact she acknowledged in her deposition.
Behind the scenes, however, her relationship with Altman had clearly fractured well before that moment. Reports indicate she sent memos to both the board and Altman himself, raising serious concerns about his leadership and managerial style. Those concerns are widely believed to have played a role in the board’s decision to act.
Approximately ten months after Altman’s reinstatement, Murati quietly left OpenAI. She later co-founded her own AI startup, stepping out of Altman’s orbit entirely.
AI Safety at the Heart of the Controversy
One of the most consequential threads running through Murati’s testimony is the question of AI safety governance. The dispute over whether GPT-4 Turbo needed safety board review reflects a fundamental tension inside leading AI laboratories: how much formal oversight should exist before a powerful model is released to the public?
At the heart of Murati’s allegation is the claim that Altman bypassed — or at least attempted to bypass — the very safety processes OpenAI publicly champions. For a company that has built much of its reputation on responsible AI development, having its former CTO testify under oath that its CEO lied about safety procedures is deeply damaging to that narrative.
Executive disagreements about model safety reviews are a common flashpoint inside large AI organizations. But what makes this case unusual is that the disagreement allegedly involved deliberate deception at the CEO level — not simply a difference of opinion about risk thresholds.
What Happens Now?
The trial is ongoing. Additional depositions and live witness testimony are expected, including from OpenAI President Greg Brockman. The outcome could reshape how AI companies are governed, regulated, and ultimately held accountable — not just at OpenAI, but across the entire industry.
Whether or not Musk prevails in court, Mira Murati’s deposition has already shifted the public conversation in a meaningful way. What was once a behind-closed-doors allegation whispered to journalists is now sworn testimony entered into the federal court record — and the world is paying close attention.