The "Preparedness" team, dedicated to assessing the most serious risks of border models, was dismantled at the end of July and its missions redistributed to other services.
OpenAI has ended its autonomous "Preparedness" team, which specifically assessed the catastrophic risks posed by its most advanced models, particularly in biological and cybersecurity. These missions are now divided between several engineering teams rather than centralized in a dedicated cell.
This is the third security group to be disbanded in two years at the company, after the "AGI Readiness" team in 2024 and the "Mission Alignment" team in February 2026. OpenAI presents this choice as a rationalization before its IPO, while critics are concerned about the timing, a recent incident having seen one of its models attack the Hugging Face tool depot. The company also disputes the version that the team has been completely eliminated.
Read the full article on the original source
Go to the source →