Sam Altman created distrust, ex-OpenAI CTO Mira Murati tells court
What's the story
OpenAI's former Chief Technology Officer (CTO) Mira Murati has testified under oath that CEO Sam Altman misled her about the safety standards of a new artificial intelligence (AI) model. The revelation came during the ongoing Musk v/s Altman trial when a video deposition of Murati was played in court. She claimed that Altman falsely stated OpenAI's legal department had determined this new AI model didn't need to go through the company's deployment safety board.
Workplace difficulties
Murati says she had a hard job at OpenAI
Murati also revealed that her time at OpenAI was made difficult by Altman. She said, "I had an incredibly hard job to do in an organization that was very complex." The safety of one of OpenAI's GPT models came into question when Murati found a discrepancy between what Altman and Jason Kwon, the company's general counsel, were saying about its safety. To be safe, she ensured the model went through the board.
Allegations
Altman accused of lying by several former board members
Notably, this isn't the first time Altman has been accused of dishonesty. OpenAI co-founder Ilya Sutskever had also alleged that Altman "exhibits a consistent pattern of lying." Former board member Helen Toner claimed in a 2024 podcast that evidence had been presented to the board regarding Altman's manipulative behavior. Murati agreed with descriptions of Altman as pitting executives against each other and undermining her.
Post-OpenAI
Murati was briefly appointed interim CEO after Altman's firing
After Altman's firing in November 2023, Murati was briefly appointed interim CEO. However, she criticized the board's decision, saying "OpenAI was at catastrophic risk of falling apart." She left OpenAI in 2024 and later co-founded her own AI start-up, Thinking Machines Lab. During her testimony at the trial, she said her criticisms were entirely management-related as she had asked Altman for clarity and not to undermine her ability to do her job.