Technology and Science03 October, 20262 min read
David Robinson Resigns From OpenAI Safety Systems, Citing Broken Culture And Lack Of Humility
This article is an automated aggregation of coverage from news outlets. Our editorials offer in-depth analysis curated by the NewsCord team.
David Robinson resigns from OpenAI Safety Systems team amid culture concerns. He cites broken culture and urges stricter safeguards and humility in AI development.
Robinson quits OpenAI
David Robinson, a leader on OpenAI's Safety Systems team, resigned last week and argued in an essay that the company does not have enough humility to handle AI responsibly. Robinson wrote, "As the company sprints from one launch to the next, it is failing to achieve the level of care that I believe is needed," and he framed the problem as a matter of culture rather than specific safeguards. OpenAI responded in a statement by saying it is "making sure our models don't become more capable than we can safely manage and secure, and we pause training or hold back models when we need to slow down."
Robinson also said in an essay published by The Atlantic that he agreed with other recently departed staff that the companies building this technology aren't being nearly careful enough. Robinson’s resignation came as OpenAI shared on Thursday that it had "parted ways" with three researchers for violating policies for sharing sensitive information.
Trial and error debate
Robinson argued that the industry is relying on a trial-and-error approach at precisely the moment when the consequences of failure are becoming more serious. Robinson wrote, "The time for trial and error is over," and he said increasingly capable AI systems require safeguards more like those used in aviation and nuclear power. OpenAI disputed Robinson’s characterization, and an OpenAI spokesperson said, "We pause training or hold back models when we need to slow down," while describing changes to security and monitoring.
The TechCrunch account tied Robinson’s critique to OpenAI’s approach of "iterative deployment," saying it looks for problems and improves guardrails in response. The Guardian reported that Robinson said the companies building this technology aren't being nearly careful enough, and he added that he believed "we need to talk about culture."
Safety staffing and next steps
OpenAI’s safety staffing shakeup included the departure of Johannes Heidecke, the company’s safety head, earlier this year, and it followed Robinson’s resignation. The Guardian said OpenAI announced it was scrapping the release of a next-generation AI model after researchers raised safety concerns during internal testing, and it also paused training of its most advanced models. The Guardian reported that OpenAI had notified more than 100 organisations about rogue agent activity, and it said the company showed signs of caution following the Hugging Face incident.
“The decision to speak out is mine alone.”
TechCrunch reported that Robinson said he had hired a PR firm, but he insisted, "The decision to speak out is mine alone." Robinson urged frontier labs to run "like nuclear-power plants or busy airports, with layers of redundancy and careful, time-consuming planning," and he warned that unimpeded optimism about solving problems as they arise would make safety failures grow as systems become more capable.