Newly-released text messages are continuing to shed light on the frenetic few days in which Sam Altman had been ousted from OpenAI.
The exchange between Altman and then-CTO Mira Murati began on November 19, 2023 — two days after the board had abruptly fired Altman as CEO, citing a loss of confidence in his leadership. The texts, terse and urgent in tone, capture a man working every lever he could find — Microsoft, Murati, even the idea of an acquisition — to claw his way back into the company he had built.
“Directionally Very Bad”
Altman opens the exchange pressing Murati for intelligence from inside the board meeting she is attending. “Can you indicate directionally good or bad?” he asks, adding that Microsoft CEO Satya Nadella and others are “anxious” for an update. Murati’s reply is blunt: “directionally very bad.”
The exchange establishes the dynamic that runs through the entire thread. Altman is on the outside, relying on Murati as his sole eyes and ears inside a board that had just ended his tenure. He is simultaneously managing Microsoft — OpenAI’s largest investor and partner — while trying to negotiate his own fate.
Board Wants Altman To Be Gone
As Murati relays that things are deteriorating, Altman pivots quickly between options. He asks if he can come in. When Murati tells him the board doesn’t want that, he offers to simply disappear: “i’m still willing to just walk away if it helps.”
It is an interesting moment — the CEO of one of the most consequential AI companies on earth offering to vanish to break a deadlock. When that gets no traction, he asks Murati to pass along that he “just wants to resolve this” and would like to join the meeting.
The board, Murati tells him, is “convinced about their decision.” When Altman asks to clarify — fired, or something new? — Murati’s response is unambiguous: “Yes for you to be gone.”
The Board Wants A New CEO That Night
What follows is a rapid-fire sequence that reveals just how far gone the board’s position was. Altman asks if he can come back to “talk about a path forward.” The board says no. He asks for more time. They say no again.
Murati then delivers the sharpest piece of news in the exchange: the board doesn’t just want Altman gone, they want a new CEO installed that same evening. “Not me,” she is careful to add.
The chosen successor turns out to be Emmett Shear, the former CEO of Twitch — an executive Altman clearly doesn’t recognise at first. “New guy is rando twitch guy,” Murati writes. “Emmett?” Altman responds. “Yeah,” she confirms.
Bringing Satya Nadella Into The Room
With the board immovable, both Altman and Murati turn to Microsoft. Murati says she is “pulling Satya now,” and Altman waits anxiously as Nadella is patched into the board call. “Is satya making progress at all?” he asks. “Satya is being diplomatic,” Murati replies — not exactly the resounding intervention Altman was hoping for.
At the same moment, a journalist — Bloomberg’s Ashlee Vance — texts Altman to say he believes the board has already lined up a new CEO. Altman forwards the screenshot to Murati. The leak from the press, arriving in real time as the board deliberated, underlines just how chaotic those 48 hours had become.
Altman Floats A Microsoft Acquisition
In what may be the most remarkable moment in the thread, Altman proposes an idea mid-conversation: “what if msft acquires openai? would that provide the governance the board wants?”
The suggestion — floated in a text message while Murati is simultaneously on a call with Satya Nadella and the OpenAI board — is a sign of how creatively Altman was scrambling. It also reflects something the board had reportedly been considering from the other direction: merging with Anthropic to keep the company out of Altman’s hands entirely.
The acquisition idea goes nowhere in the texts. But it would not be the last time the board’s governance failures would raise questions about OpenAI’s corporate structure.
“They Don’t Care If Everyone Quits”
Perhaps the most chilling line in the thread comes when Altman asks whether the team should send a letter to the board. Murati’s response is flat: “It won’t matter. They don’t care if everyone quits.”
It is a remarkable admission — that the board was prepared to see the entire company walk out the door rather than reverse course. Altman responds with a dark theory: “is that what they want is the ip going to anthropic?” His team, he says, thinks that is the endgame.
Murati’s correction is sharp: “Just not your hand on agi.”
The line is emblematic of what the coup was really about. This wasn’t a standard corporate governance dispute. It was a fight over who would control the path to artificial general intelligence — a prize the board members believed was too important to be left in Altman’s hands.
The Reversal
What the texts don’t capture is the hours that followed. The employee revolt that Murati dismissed as ineffective turned out to be exactly what broke the board. More than 700 OpenAI employees signed a letter threatening to resign and follow Altman to Microsoft. Even Ilya Sutskever, the chief scientist who had orchestrated the ouster, signed it. Within five days, Altman was back as CEO.
Murati herself left OpenAI in late 2024 to found her own AI company, Thinking Machines. The board members who ousted Altman were replaced. And the coup — for all its drama — changed almost nothing, except perhaps hardening Altman’s grip on the company.
Full Transcript
Sam Altman: can you indicate directionally good or bad? satya and others anxious
Mira Murati: directionally very bad
Sam Altman: ok
Sam Altman: can you wrap up soon? lots of pressure from msft for an update
Mira: Sam this is very bad
Sam: can i come in?
Mira: They don’t want you to
Sam: what do you want to make it better? i’m still willing to just walk away if it helps
Sam: if they’re ramped up for crazy lawsuits against me then i’m not sure what
Sam: can you please tell them i just want to resolve this however and would like to join
Mira: They’re convinced about their decision
Sam: for me to be fired? or some new thing?
Mira: Yes for you to be gone
Sam: ok
Sam: then can i come back and talk about a path forward with them?
Mira: They’re saying no and they need more time
Sam: more time for what?
Mira: They’ve walked me through all the reasons and the issues with you and why you can’t be ceo
Sam: can you ask why they’ve been a saying all weekend they wanted me back?
Mira: They want new ceo in place
Sam: can you say you will call back in 10 min
Mira: They want to have a new ceo in place tonight (not me
Sam: do they know who?
Sam: can i tell satya? is this final?
Sam: or, should i add satya in?
Mira: Trying to add Satya now
Sam: still don’t want me?
Mira: New guy is rando twitch guy
Mira: They don’t want you
Sam: emmett?
Mira: Yeah
Mira: But hold on I’m pulling Satya now
Sam: ok
Sam: do you think any way you can turn this around? even if we let them sleep on it tonight or whatever?
Mira: Have been trying
Sam: from a journalist
Mira: Yeah
Mira: Hoping Satya can help undo this
Sam: should team send letter to board now?
Sam: is he on with them?
Mira: It won’t matter
Mira: They don’t care if everyone quits
Sam: is that what they want is the ip going to anthropic?
Sam: that’s what team thinks
Mira: Just not your hand on agi
Sam: i can not come back!
Sam: they were the ones that asked as of yesterday morning
Sam: did satya get on call?
Mira: Yes with him
Sam: wait i have an interesting idea
Mira: Still with Satya
Mira: Go ahead
Sam: what if msft acquires openai? would that provide the governance the board wants?
Sam: is satya making progress at all?
Mira: Satya is being diplomatic
Sam: will you emmett with the team asap?
Sam: also does adam know you rehired me?
Mira: Yes