Brief IA

Mira Murati Accuses Sam Altman of Lies About OpenAI's Safety

🤖 Models & LLM·Tom Levy·

Mira Murati Accuses Sam Altman of Lies About OpenAI's Safety

Mira Murati Accuses Sam Altman of Lies About OpenAI's Safety
Key Takeaways
1Mira Murati, former CTO of OpenAI, claims that Sam Altman lied about the safety standards of an AI model.
2She discovered a disagreement between Altman and legal advisor Jason Kwon regarding the need to go through the safety committee.
3The OpenAI board dismissed Altman for lack of transparency, according to Murati and other testimonies.
💡Why it mattersThese accusations raise questions about governance and transparency within AI companies.
Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄
Full Analysis

Mira Murati Questions Sam Altman's Integrity

During a sworn testimony, Mira Murati, former Chief Technology Officer of OpenAI, stated that Sam Altman, the company's CEO, had lied to her about the safety standards of a new artificial intelligence model. This testimony was presented during the ongoing trial between Elon Musk and Altman. Murati claimed that Altman falsely indicated that OpenAI's legal department had concluded that a new AI model did not need approval from the company's deployment safety committee. When asked about the truthfulness of Altman's statements, Murati unequivocally replied, "No."

Controversial Management at OpenAI

Murati described the challenges she faced during her tenure at OpenAI, emphasizing that Altman's management complicated her work. She expressed the need for clear leadership and criticized Altman for undermining her ability to perform her tasks. A notable example concerns the safety of one of OpenAI's GPT models. After discussing with Altman, Murati consulted Jason Kwon, OpenAI's legal advisor since 2021, and found a disagreement between Kwon's statements and Altman's. As a precaution, she insisted that the model be reviewed by the safety committee.

Recurring Accusations of Lying

Sam Altman is no stranger to accusations of lying. Ilya Sutskever, co-founder of OpenAI, wrote a 52-page memo for the board, claiming that Altman had a "consistent pattern of lying" and that he pitted executives against each other. Helen Toner, a former board member, also mentioned evidence of manipulation by Altman in a 2024 podcast. Murati confirmed these descriptions, highlighting that Altman undermined her authority.

Consequences for OpenAI's Leadership

The board of directors dismissed Altman, stating that he was not always forthcoming in his communications, which hindered their responsibilities. When asked why the board had lost confidence in him, Altman replied that it was a question to ask the board. After Altman's dismissal, Murati was briefly appointed interim CEO, but she criticized this decision, stating that OpenAI was at risk of collapsing. She left the company in 2024 to found Thinking Machines Lab, a competitor to OpenAI.

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.