The Deposition That Exposed OpenAI’s Internal Crisis
Newly released court documents from the Musk v. Altman lawsuit reveal shocking details about OpenAI’s near-collapse in November 2023. A 10-hour sworn deposition from co-founder Ilya Sutskever paints a dramatic picture of panic, unverified allegations, and catastrophic governance failures that nearly destroyed the AI giant.
Key Revelations from the Testimony
The deposition transcript, released this week, shows how brilliant AI researchers made disastrous organizational decisions, with some board members preferring to destroy OpenAI rather than let it continue under Sam Altman’s leadership. The testimony centers on Elon Musk’s claim that OpenAI betrayed its original nonprofit mission by aligning with Microsoft.
The 52-Page Dossier Against Altman
Sutskever prepared an extensive 52-page case for removing Altman, complete with screenshots and organized documentation. He testified that he explicitly stated in the memo: “Sam exhibits a consistent pattern of lying, undermining his execs, and pitting his execs against one another.” The full brief was sent using disappearing email technology due to concerns about leaks and has not been publicly released.
Year-Long Board Politics
When asked how long he’d been considering firing Altman, Sutskever answered: “At least a year.” He was waiting for favorable board dynamics, specifically “that the majority of the board is not obviously friendly with Sam.” This reveals a calculated, long-term political strategy despite the public appearance of close collaboration between the two leaders.
The Weekend OpenAI Almost Ceased to Exist
Within 48 hours of Altman’s firing, active discussions began about merging OpenAI with competitor Anthropic. Former board member Helen Toner was “the most supportive” of this direction, which would have eliminated OpenAI as an independent entity. Sutskever testified he was “very unhappy about it” and “really did not want OpenAI to merge with Anthropic.”
Ideological Divide and Mission Interpretation
The deposition reveals deep ideological divides within OpenAI’s leadership. When executives warned that the company would collapse without Altman, Toner responded that destroying OpenAI could be consistent with its safety mission. This represents a strand of AI safety thinking that views rapid AI development as potentially more dangerous than no development at all.
Critical Miscalculations and Governance Failures
Sutskever admitted that nearly everything in his 52-page memo came from one source: CTO Mira Murati. He didn’t verify claims with other executives and trusted Murati completely. When asked about the board’s process, he was blunt: “One thing I can say is that the process was rushed. I think it was rushed because the board was inexperienced.”
Workforce Loyalty Surprise
The board fundamentally miscalculated employee loyalty. When 700 of 770 employees signed a letter demanding Altman’s return and threatening to leave for Microsoft, Sutskever was genuinely surprised. He testified: “I had not expected them to cheer, but I had not expected them to feel strongly either way.”
Lessons in AI Governance and Leadership
The deposition serves as a cautionary tale about what happens when technical expertise meets organizational incompetence. It highlights the critical importance of proper governance structures, verification processes, and understanding organizational dynamics in high-stakes technology companies. The case continues to raise fundamental questions about who controls advanced AI models and whether they can be developed safely in the public interest.
