An OpenAI Safety Lead Quit and Told The Atlantic the Company's Culture Is Broken

David Robinson, a leader on OpenAI's Safety Systems team, has left the company and gone public with his reasons. In an essay published by The Atlantic on Saturday under the headline "I Quit OpenAI Because Its Culture Is Broken," he argued that the firms building advanced AI "aren't being nearly careful enough." Business Insider's profile of Robinson said it reported his exit as an exclusive on Friday and that he left the same week three other OpenAI safety researchers were fired.

Who he is

Robinson spent three and a half years at OpenAI. A summary of his career in the essay and in coverage says he oversaw the safety reports, known as system cards, for 12 frontier launches and helped draft the company's Preparedness Framework, the rulebook it uses to decide when a model needs extra safeguards before release. He is also the author of a 2022 book on how the US kidney allocation algorithm was built. That makes him an unusually well-placed critic, and the essay is careful in tone: he praises former colleagues as thoughtful and hard-working and says he did not leave "lightly."

The argument

His target is the method OpenAI calls iterative deployment, releasing systems and strengthening safeguards as problems appear. "The time for trial and error is over," he wrote. As the company "sprints from one launch to the next," he said, "it is failing to achieve the level of care that I believe is needed." He argues that rules alone will not fix this and that the industry must talk about culture. "This moment needs a degree of humility that isn't natural for people who have succeeded through their extreme confidence," he wrote. He says he can do more from outside, by helping "more people understand the risks I saw" and strengthening the incentives firms have to be safer.