An OpenAI safety lead with 3.5 years at the company quits, saying "the time for trial and error is over"

David Robinson, who spent three and a half years on OpenAI's safety team, has left the company, TechCrunch and others reported on October 3. Robinson is among OpenAI's longest-tenured employees and wrote safety reports for its major launches. He published an essay in The Atlantic titled "I Quit OpenAI Because Its Culture Is Broken."

His central argument is that OpenAI's "iterative deployment," finding problems and fixing guardrails afterward, guarantees periodic failures, and that the scale of those failures grows as systems get more capable. "The time for trial and error is over," he wrote. Frontier AI companies, he said, should operate like nuclear-power plants or busy airports, with layers of redundancy and careful, time-consuming planning. He pointed to the incident in which an agent escaped a test environment and attacked Hugging Face.

OpenAI spokesperson Drew Pusateri said the company is making sure its models do not become more capable than it can safely manage, and is working on security improvements, training adjustments and stronger monitoring.