To minimize these risks as ai models continue to improve, we are creating a new team called Preparedness. Led by Aleksander Madry, the Readiness team will closely connect capabilities assessment, assessments and internal red teaming for frontier models, from the models we will develop in the near future to those with AGI level capabilities. The team will help track, assess, forecast and protect against catastrophic risks spanning multiple categories, including:
- Individualized persuasion
- Cyber security
- Chemical, biological, radiological and nuclear (CBRN) threats
- Autonomous Replication and Adaptation (ARA)
The readiness team’s mission also includes the development and maintenance of a Risk-Based Development Policy (RDP). Our PDR will detail our approach to developing rigorous assessments and monitoring of border model capacity, creating a spectrum of protective actions, and establishing a governance structure for accountability and oversight across that development process. The PDR is intended to complement and expand our existing risk mitigation work, contributing to the security and alignment of new, high-capacity systems, both before and after deployment.
