When Sam Altman was suddenly sacked as CEO of OpenAI in November 2023, before being reinstated days later following a dramatic employee revolt, the tech world was left reeling.
At the time, the board said only that Altman had not been “consistently candid,”a cryptic phrase which raised more questions than answers. Now, explosive testimony from Ilya Sutskever, the co-founder and former chief scientist of OpenAI, has pulled back the curtain on what really drove one of Silicon Valley’s most dramatic boardroom battles.
Sutskever’s deposition, part of Elon Musk’s ongoing lawsuit against Altman and OpenAI, is a portrait of a CEO who worked, in large part, by manipulation and strategic deception. During nearly 10 hours of questioning, Sutskever painted a picture of a leader who told different people different things and created confusion and conflict among his top executives.
Altman’s ‘Chameleon-Like Ability’ & Undermining of Execs, Sutskever Details Grievances to OpenAI Board
As Altman’s friend Sutskever testified, he had a tendency to tell people exactly what they wanted to hear, even if it wasn’t what he was actually intending. This chameleon-like ability let him pursue his own agenda, said Sutskever, while executives were kept off balance, unable to determine the company’s true direction.
The concerns weren’t new or sudden. Sutskever said he’d been thinking about removing Altman for “at least a year” before actually firing him in November 2023, which implies that deep-seated and systemic issues were at play, rather than the result of any one incident.
What makes Sutskever’s testimony particularly significant is that he wasn’t just sharing opinions-he had been systematically documenting concerns. At the request of one or more of OpenAI’s independent board members, Sutskever compiled a 52-page memo filled with screenshots and other evidence detailing executives’ grievances about Altman’s leadership style.
The memo went to the board members Adam D’Angelo, Helen Toner, and Tasha McCauley. Sutskever also prepared a separate document raising concerns about OpenAI president Greg Brockman. Both memos were sent as disappearing emails, though copies remain in the hands of “various lawyers,” according to the deposition.
Sutskever’s core allegation was stark: Altman “exhibits a consistent pattern of lying, undermining his execs, and pitting his execs against one another.” He supplied specific examples to back up the claim.
Sutskever’s Testimony Details Alleged Internal Conflicts and Altman’s Ambiguous Tactics at OpenAI
One such example was with Sutskever and Jakub Pachocki, who is now the chief scientist at OpenAI. Sutskever says Altman told each of the two “conflicting things about the way the company would be run,” a move that pitted two key research leaders against one another.
Another example concerned the fate of OpenAI’s former research executive, Dario Amodei, who wanted to oversee all research at the company and sought to have Brockman removed. Instead of taking a clear stance, Sutskever said Altman remained strategically ambiguous-appearing to wait until it became clear which faction would emerge stronger before committing his support.
The testimony also addressed issues highlighted by Mira Murati, OpenAI’s former chief technology officer. Sutskever testified that Murati accused Altman of introducing friction in her relationship with Daniela Amodei, a research executive then at OpenAI. Murati reportedly shared with Sutskever screenshots and documentation, including text messages between Altman and Brockman.
Sutskever added that even his concerns about Altman’s leadership style predated OpenAI. Similar behavioral issues, according to his testimony, were cited when Altman left his leadership position at Y Combinator, the prestigious startup accelerator.
‘Events of 2023 Are Behind Us’ as Sutskever Testimony Adds Context
OpenAI has pushed back firmly against these allegations. A company spokesperson issued the following statement: “The events of 2023 are behind us. These claims were fully examined during the board’s independent review, which unanimously concluded Sam and Greg are the right leaders for OpenAI.”
This response echoes comments made in 2024 by board chair Bret Taylor following the company’s internal investigation into the matter.
One thing is certain: The temporary removal and swift return of Altman marked a turning point for OpenAI. That hundreds of employees would threaten to quit rather than work under anyone but Altman suggests a deep well of support at the lower levels of the company, even if his management style had top executives reportedly deeply concerned.
These new revelations from the Sutskever deposition add important context to an episode that has continued to shape OpenAI’s trajectory and raises questions about corporate governance in the fast-changing world of AI.
