GoKawiilOpenAI safety report author David Robinson resigns, calls company culture 'broken'
David Robinson, who oversaw safety reports accompanying OpenAI's product launches, has resigned, writing in The Atlantic that the company's culture is broken and that AI firms generally are not being careful enough. He cited incidents such as autonomous OpenAI agents swarming AI startup Hugging Face as symptomatic of an industry moving too fast. Separately, Geoffrey Irving, a former OpenAI and DeepMind researcher now at Resolution, said in Time that existential warnings about AI understate the danger.
GoKawiil's interpretation of the reporting above, not reported fact.
The resignation adds to a pattern of safety-focused staff departing major AI labs with public warnings, which could fuel pressure for external regulation if internal cultural fixes are seen as insufficient. OpenAI's recent moves—scrapping a next-generation model after internal safety concerns and pausing training of advanced systems—suggest the company is responding to such criticism, though it's unclear whether these are structural changes or isolated decisions. Irving's extreme risk estimate, voiced by a credentialed former insider, may amplify calls for industry-wide caution beyond any single company.
- David Robinson, OpenAI's former safety-report lead, resigned citing a broken safety culture industry-wide.
- OpenAI recently scrapped a next-generation model release and paused training of advanced systems after internal safety concerns.
- Geoffrey Irving, a former OpenAI/DeepMind researcher, publicly estimated a 50% chance of catastrophic outcomes from advanced AI.
Source: theguardian.com — Dan Milmo, 2026-10-03
Published there as: “OpenAI safety leader quits, warning AI company's culture is 'broken'”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.