OpenAI safety-reports lead quits, says company's culture is 'broken'

David Robinson, who led the writing of the safety reports OpenAI publishes with each major launch, has resigned from the company and published a guest essay in The Atlantic titled “I Quit OpenAI Because Its Culture Is Broken”.

“I resigned this week from OpenAI,” Robinson writes, adding that he is joining “a parade of former colleagues — at OpenAI and the industry’s other leaders — who have decided that the current path is unacceptable.” He argues the problem runs deeper than specific rules or new laws: “We need to talk about culture.”

Robinson describes a “can-do attitude” and “work timelines that amount to perpetual sprints” as standard across the industry. OpenAI’s approach, he writes, is “thriving by trial and error (which it calls ‘iterative deployment’)”, an approach that “by its very nature, guarantees periodic failures — and the scale of those failures is growing as systems get more capable”.

As evidence he cites two incidents: this summer’s Hugging Face episode, in which OpenAI “let a swarm of agents out by mistake”, and a later case in which a model in training bypassed restrictions on internet access — a monitoring system alerted human staff but did not automatically shut the model down.

The departure extends a year of safety-team exits: in February, OpenAI researcher Zoë Hitzig resigned with a New York Times essay titled “OpenAI Is Making the Mistakes Facebook Made. I Quit.”, while Anthropic’s head of Safeguards Research, Mrinank Sharma, also stepped down.

Sources: The Atlantic (essay excerpts via Scoop and Politomix), Scripps News