Murati testifies Altman ‘sowed chaos’ as Musk OpenAI trial enters week two

TL;DR:

• Former OpenAI CTO Mira Murati testified by recorded video that CEO Sam Altman was “creating chaos” and could be deceptive with executives, and that the company was at “catastrophic risk of falling apart” during the November 2023 board crisis.
• Ex-board member Shivon Zilis told the court the board “voiced extreme concern” about the release of ChatGPT “without any semblance of board communication”; she now works for Elon Musk’s Neuralink and is the mother of four of Musk’s children.
• Resultsense view: the depositions add unusually direct insider colour to a question UK enterprise buyers care about — the operational maturity of frontier-lab governance — at the same week Anthropic announces a $50 billion infrastructure programme and OpenAI publishes B2B research about “frontier firm” enterprise discipline.

The testimony was played in Oakland, California, federal court in the second week of trial in Elon Musk’s lawsuit against OpenAI. Musk, an OpenAI co-founder, sued the company in 2024 alleging it improperly converted to a for-profit structure and abandoned charitable goals. He is seeking $150 billion in damages from OpenAI and investor Microsoft, to be paid to the company’s charitable arm.

What Murati told the court

“My concern was about Sam saying one thing to one person and completely the opposite to another person,” Murati said. She testified Altman pitted executives against one another and undermined her role as technology chief. Notably, Murati nonetheless said she had wanted Altman to remain CEO and had pressed the board for fuller justification before the brief 2023 ouster. She has since left OpenAI to co-found her own AI start-up. The throughline of her testimony was operational risk: “OpenAI was at catastrophic risk of falling apart. I was concerned about the company completely blowing up.”

Zilis on the ChatGPT release

Zilis described the board as having “voiced extreme concern” about the chatbot’s release “without any semblance of board communication”, and confirmed she had separately raised concerns about Altman internally. The detail is significant because the launch of ChatGPT 3.5 in late 2022 is what triggered the entire current frontier-lab arms race.

Why this matters beyond the courtroom

If Musk succeeds, the implications go beyond the headline damages number. He could benefit competitively — xAI is now part of SpaceX, which separately signed a major compute supply deal with Anthropic this week — and the structural restoration of OpenAI’s nonprofit posture would reshape the commercial ambitions of the most-watched AI company. UK procurement teams negotiating multi-year frontier-AI contracts have every reason to care about the corporate-form question.

Looking forward

The trial has already produced surprises, including reporting that Musk attempted to settle with OpenAI president Greg Brockman days before proceedings began. The testimony’s value is largely retrospective; the live procurement question is whether the dispute changes Microsoft’s posture, given the $150 billion figure aimed at it, and how quickly the trial concludes ahead of OpenAI’s next major commercial milestone.