Chaos and lies: Why Sam Altman was booted from OpenAI according to new testimony (www.theverge.com)

New testimony from OpenAI cofounder Ilya Sutskever, deposed in Elon Musk’s lawsuit against Sam Altman and OpenAI, paints a detailed picture of the internal fractures that led to Altman’s brief November 2023 ouster. Sutskever says he compiled a 52‑page memo, screenshots and disappearing emails showing Altman repeatedly gave conflicting accounts to senior staff, “pit[ted]” executives against one another, and withheld or misrepresented information — behavior he and some board members felt made him untrustworthy to lead a company pursuing AGI. He also described short-lived merger discussions with Anthropic during the chaos and confirmed memos about Greg Brockman and other executives were circulated to independent board members (and are now in lawyers’ hands). Sutskever acknowledged some of his claims were based on secondhand reports and stressed the importance of firsthand evidence. For the AI community the episode underscores that governance, transparency and accurate safety reporting are as consequential as technical progress. The dispute exposed alleged gaps in OpenAI’s formal safety processes and potential conflicts (e.g., a disclosed startup fund), prompted executive departures and spawned rival startups (Safe Superintelligence, Thinking Machines), and shows how leadership behavior can shape research priorities and public trust. OpenAI says an independent review cleared Altman and Greg Brockman to lead, but ongoing litigation and further depositions could reveal more about how oversight failures interact with the development and deployment of powerful AI.