OpenAI disbanded its Preparedness team for catastrophic AI risk, reassigning the work

OpenAI shut down its Preparedness team, which assessed whether its own models could pose catastrophic risks, and folded its work into other groups as safety staff have departed.