Court filings unseal OpenAI internal chats revealing board disputes

Court filings in the Musk v. OpenAI trial have done something unusual — they turned years of private Slack-style chatter, texts, emails, and deposition clips into a public map of how OpenAI actually worked behind the scenes. And the picture is messy. You see board members distrusting management, executives distrusting each other, Microsoft worrying about the company early on, and safety arguments colliding with product pressure. That matters because this is not just gossip from a famous startup. It is evidence in a case that could shape how OpenAI is governed and how much freedom AI labs have to mix nonprofit ideals with very commercial behavior. (nytimes.com) ### What got unsealed? The trial has surfaced internal documents from multiple eras of OpenAI’s life — early partnership talks with Microsoft, the November 2023 board coup that briefly ousted Sam Altman, and recent testimony from former executives. The documents include texts, emails, diary-style notes, and sworn testimony introduced as evidence in Musk’s suit, which argues that OpenAI abandoned its original nonprofit mission as it became more commercial. (theverge.com) ### Why are people focused on Mira Murati? Murati’s testimony is the clearest inside account of the leadership breakdown. She said Altman created distrust and chaos among top executives, and described a pattern of pitting leaders against each other. That matters because Murati was not an outside critic — she was OpenAI’s CTO and briefly interim CEO during the 2023 firing saga, so her account goes straight to the que(theverge.com). (usnews.com) ### What did the texts show? The 2023 messages around Altman’s ouster made the board fight feel very immediate. Business Insider highlighted exchanges in which Murati told Altman the board did not want him back, even as employees and Microsoft pushed for his return. Basically, the texts show how quickly a corporate crisis stops sounding stra(usnews.com)jury to read them later. (businessinsider.com) ### Why do the Microsoft emails matter? One of the most revealing batches came from 2018, before Microsoft fully committed to OpenAI. Wired reported that Microsoft executives were skeptical of OpenAI and of Altman, but still worried that backing away could drive the startup toward Amazon instead. So the alliance that later defined the AI boom did not begin with perfect trust. It began with strategic anxiety, competitive fear, and a bet that controlling the relationship mattered more than loving it. (wired.com) ### Is this really about safety? Partly, yes — but not in the clean, abstract way AI companies like to discuss safety in blog posts. The trial has pulled safety into the same frame as power struggles, board process, and product speed. The tension running through the evidence is simple: if leaders do not trust each other, then promises about careful deployment and mission-first governance get harder to believe. That is why the filings have drawn attention well beyond the specific Musk lawsuit. (msn.com) ### What does Musk actually want? Musk says OpenAI broke its founding deal by shifting from a nonprofit mission toward profit and closed commercial advantage. OpenAI says Musk’s claims distort both the company’s structure and his own role in trying to reshape it years ago. The trial is not just about old grudges. It is about whether the legal system will treat OpenAI’s original mission statements as binding commitments or as startup-era aspirations that later got overtaken by reality. (theverge.com) ### Why does this land so hard now? Because OpenAI is no longer a quirky lab with internal drama. It sits near the center of the AI economy, with Microsoft deeply tied to its products and rivals building strategy around its moves. So every newly unsealed message does two jobs at once — it feeds the courtroom fight, and it forces investors, regulators, and the public to ask whether the company’s internal controls were ever strong enough for the power it now holds. (cnbc.com) ### Bottom line? The big reveal is not that OpenAI had conflict — every fast-growing company does. It is that the conflict touched the exact things OpenAI says make it different: mission, safety, and governance. Once those arguments are preserved in texts and entered into evidence, they stop being internal culture problems and start becoming part of the company’s public record.