URGENT: OpenAI has apparently disbanded its “Preparedness” team, which was responsible for assessing catastrophic risks of its AI models.

The team’s work in biosafety and cybersecurity is being reassigned to existing groups.