Inside the Slack receipts that got Sam Altman fired

The dust may have settled on OpenAI's 2023 boardroom coup, but the real receipts are just starting to surface.

An excerpt from the upcoming book "The Optimist: Sam Altman, OpenAI, and the Race to Invent the Future," by Wall Street Journal reporter Keach Hagey, reveals fresh details about why the board pulled the trigger on Sam Altman.

Slack receipts and paper trails

According to the book, the board became increasingly alarmed after discovering issues like the OpenAI Startup Fund being legally owned by Altman himself rather than the entity.

At the same time, co-founder Ilya Sutskever and CTO Mira Murati were actively gathering evidence of what they viewed as Altman's toxic and dishonest behavior, complete with screenshots from Murati's Slack.

One glaring example involved GPT-4 Turbo. Altman allegedly told others that OpenAI's legal team said the model didn't need to be reviewed by the joint safety board. The company's top lawyer later denied ever saying that.

Sutskever took this evidence to the board, prompting Altman's firing and Murati's brief stint as interim CEO. We all know how that ended: employees revolted, Altman returned within days, and both Sutskever and Murati eventually left to build their own AI startups.

Why it matters

These details paint a much messier picture than just a philosophical clash over AI safety: it was a full-blown breakdown of trust and governance inside Silicon Valley's most important startup.

It shows that internal leadership was genuinely spooked by Altman allegedly sidestepping safety guardrails and fudging legal advice. Yet, despite having actual receipts, the board still failed completely against Altman's sheer operational leverage and employee loyalty.

We think this explains why OpenAI's old non-profit governance model was doomed from the start. When the founder holds all the cultural and commercial chips, a board without real enforcement power is just performing theater.