As OpenAI’s board moved to remove Sam Altman in November 2023, the company’s top executives were exchanging frantic messages about leadership at OpenAI, Microsoft, and its CEO, Satya Nadella, among others.
Newly released chats between Altman and Mira Murati that were disclosed in the matter of Musk v. Altman (2026), and shared by the X account Internal Tech Emails, showcase the confusion and panic inside OpenAI during the 48 hours that nearly changed the company’s future.
(Musk v. Altman is a high-stakes, April 2026, federal trial in which Tesla chief Elon Musk accuses Altman of betraying OpenAI’s founding mission as a nonprofit.)
Murati at one stage even told Altman that the board “don’t care if everyone quits”.
The texts show Altman trying every possible route back into OpenAI while Murati relayed updates from a board determined to push him out.
“Directionally very bad”
The exchange is from November 19, 2023, shortly after OpenAI’s board fired Altman, citing a breakdown in trust.
Altman repeatedly asked Murati for signals on how the board discussions were unfolding. “Can you indicate directionally good or bad?” he wrote, noting that Nadella and others were waiting ‘anxiously’ for updates.
Murati’s answer was immediate: “directionally very bad.”
The conversation quickly made clear that the board was not reconsidering its position. Altman asked whether he could join the discussions or help find “a path forward”. Murati told him the board did not want him involved.
At one point, Altman even suggested stepping away entirely, writing: “I’m still willing to just walk away if it helps.”
A replacement lined up
The board was already moving towards a replacement.
Murati informed Altman that Emmett Shear, the former chief executive of Twitch, was being lined up to take over as interim CEO that same evening.
“New guy is rando twitch guy,” Murati wrote. “Emmett?” Altman replied.
“Yeah,” she answered.
Microsoft enters the picture
As tensions rose, both executives looked towards Microsoft for support.
Murati told Altman she was “pulling Satya now” as Nadella joined conversations with the board. But the intervention appeared limited.
“is satya making progress at all?” Altman later asked.
“Satya is being diplomatic,” Murati replied.
Altman then floated the possibility of Microsoft acquiring OpenAI altogether.
“what if msft acquires openai? would that provide the governance the board wants?” he asked.
The idea did not appear to gain traction.
“Just not your hand on agi”
The darkest exchange came when Altman suggested OpenAI employees send a letter pressuring the board to reverse course.
Murati’s response was stark: “It won’t matter. They don’t care if everyone quits.”
Altman then questioned whether the board wanted OpenAI’s intellectual property to end up with rival Anthropic.
Murati replied: “Just not your hand on agi.”
Days later, the board’s position collapsed after hundreds of OpenAI employees threatened to resign and join Microsoft alongside Altman. He returned as CEO within a week.
The texts resurfaced as Murati testified this week in Musk’s lawsuit against Altman, where she accused him of “creating chaos” inside OpenAI.
"My concern was about Sam saying one thing to one person and completely the opposite to another person," said Murati, according to a report by Reuters. She said Altman was at times deceptive with her and others.
Newly released chats between Altman and Mira Murati that were disclosed in the matter of Musk v. Altman (2026), and shared by the X account Internal Tech Emails, showcase the confusion and panic inside OpenAI during the 48 hours that nearly changed the company’s future.
(Musk v. Altman is a high-stakes, April 2026, federal trial in which Tesla chief Elon Musk accuses Altman of betraying OpenAI’s founding mission as a nonprofit.)
Murati at one stage even told Altman that the board “don’t care if everyone quits”.
The texts show Altman trying every possible route back into OpenAI while Murati relayed updates from a board determined to push him out.
“Directionally very bad”
The exchange is from November 19, 2023, shortly after OpenAI’s board fired Altman, citing a breakdown in trust.
Altman repeatedly asked Murati for signals on how the board discussions were unfolding. “Can you indicate directionally good or bad?” he wrote, noting that Nadella and others were waiting ‘anxiously’ for updates.
Murati’s answer was immediate: “directionally very bad.”
The conversation quickly made clear that the board was not reconsidering its position. Altman asked whether he could join the discussions or help find “a path forward”. Murati told him the board did not want him involved.
At one point, Altman even suggested stepping away entirely, writing: “I’m still willing to just walk away if it helps.”
A replacement lined up
The board was already moving towards a replacement.
Murati informed Altman that Emmett Shear, the former chief executive of Twitch, was being lined up to take over as interim CEO that same evening.
“New guy is rando twitch guy,” Murati wrote. “Emmett?” Altman replied.
“Yeah,” she answered.
Microsoft enters the picture
As tensions rose, both executives looked towards Microsoft for support.
Murati told Altman she was “pulling Satya now” as Nadella joined conversations with the board. But the intervention appeared limited.
“is satya making progress at all?” Altman later asked.
“Satya is being diplomatic,” Murati replied.
Altman then floated the possibility of Microsoft acquiring OpenAI altogether.
“what if msft acquires openai? would that provide the governance the board wants?” he asked.
The idea did not appear to gain traction.
“Just not your hand on agi”
The darkest exchange came when Altman suggested OpenAI employees send a letter pressuring the board to reverse course.
Murati’s response was stark: “It won’t matter. They don’t care if everyone quits.”
Altman then questioned whether the board wanted OpenAI’s intellectual property to end up with rival Anthropic.
Murati replied: “Just not your hand on agi.”
Days later, the board’s position collapsed after hundreds of OpenAI employees threatened to resign and join Microsoft alongside Altman. He returned as CEO within a week.
The texts resurfaced as Murati testified this week in Musk’s lawsuit against Altman, where she accused him of “creating chaos” inside OpenAI.
"My concern was about Sam saying one thing to one person and completely the opposite to another person," said Murati, according to a report by Reuters. She said Altman was at times deceptive with her and others.
