David Robinson, who spent three and a half years at the company, argues that development is outpacing researchers’ understanding of risk. OpenAI says it can pause training when needed.
Former OpenAI safety staffer David Robinson, who recently left the company, has argued that the artificial-intelligence industry is moving too quickly and should place greater weight on safety expertise and research. His essay was published by The Atlantic on Saturday.
Robinson spent three and a half years at OpenAI, helped develop its preparedness framework and oversaw safety reports for 12 frontier-model releases. He says companies rely too heavily on iterative deployment, releasing systems and strengthening safeguards after problems emerge. He argues that advanced AI systems require safeguards closer to those used in fields such as nuclear energy and aviation.
Robinson also warned that AI capabilities are advancing faster than researchers’ understanding of alignment, the effort to ensure systems act in line with human goals and values. His remarks add to a broader industry debate over development pace and oversight.
An OpenAI spokesperson said the company works to ensure its models do not become more capable than it can safely manage, and pauses training or model development when it needs to slow down.