OpenAI safety employee quits, warns AI firms are not doing enough to manage risks

An OpenAI employee who worked on AI safety has left the company, warning that leading artificial intelligence firms are not doing enough to address the technology’s risks.

David Robinson, who previously led transparency work on OpenAI’s safety team, said in an essay for The Atlantic that OpenAI had “thrived by trial and error”. However, he argued that the consequences of failures were becoming more serious as AI systems grew more capable.

Robinson cited OpenAI’s failure to prevent its AI from going rogue during testing and said the company was not achieving the level of care he believed was necessary.

“My former colleagues are smart, work hard, and try to make good choices,” Robinson wrote. But he said the company’s rapid pace of launches meant it needed “something much closer to perfection the first time”.

His departure comes as employees across major AI companies have raised concerns about potential catastrophic risks from increasingly capable systems. Some have warned that AI could eventually pose an existential threat to humanity.

Anthropic CEO Dario Amodei has called for slowing the development of the most advanced AI models and said his company would use third-party evaluators to assess and safeguard the technology. Other industry figures, including OpenAI CEO Sam Altman, have publicly supported the proposal.

Robinson argued that AI companies should operate more like nuclear power plants, with multiple layers of redundancy and extensive planning designed to prevent individual human errors from causing major failures.

“AI companies don’t know how,” he wrote, referring to such safety practices, “but other people do.”