OpenAI safety leader quits, calls company culture 'broken'

OpenAI has lost the safety leader who wrote the safety reports accompanying its product releases, and he left with a public accusation: the company's culture is "broken" and AI firms are not "being nearly careful enough" about developing the technology.

David Robinson explained his resignation in an essay headlined, "I quit OpenAI because its culture is broken." He led the writing of safety reports that accompanied the ChatGPT developer's product releases, making him one of the most visible internal voices on how OpenAI assessed and disclosed the risks of its own models.

His departure puts a named, senior safety figure on the record with a direct critique of the company behind ChatGPT, at a moment when regulators, enterprise customers and investors are pressing AI developers to demonstrate that safety governance is more than a compliance exercise.

Who is David Robinson and what did he do at OpenAI?

Robinson led the writing of the safety reports that accompanied OpenAI's product releases. Those documents served as the company's formal account of how it evaluated risks before shipping products built on its models, including ChatGPT.

His role placed him at the intersection of engineering and accountability: translating internal risk assessments into published reporting that outside observers relied on to judge how carefully OpenAI was handling its technology.

What does his resignation essay say?

Robinson did not resign quietly. He published an essay under the headline, "I quit OpenAI because its culture is broken," and in it he warned that AI firms were not "being nearly careful enough" about developing the technology.

The warning was aimed beyond his own employer. He joins other insiders in urging the industry to take more care over rapidly developing technology, according to the report.

Why does one departure matter for the AI industry?

Safety leadership departures carry weight because they test whether a company's stated commitment to responsible development survives contact with commercial pressure. When the person responsible for documenting product safety concludes that the underlying culture is broken, customers and partners must ask what those safety reports actually reflected.

Robinson is not an isolated voice. He joins other insiders who have urged the AI industry to take more care as the technology develops rapidly — a pattern that keeps the spotlight on how OpenAI and its rivals balance shipping speed against risk management.

The essay, and the industry-wide warning inside it, now sits in public view. How OpenAI responds to the charge that its culture is broken — and whether its future safety reporting retains credibility with the insiders who produce it — will shape how the market reads the company's next product release.