METR has detailed its comprehensive approach to safeguarding confidential information and non-public AI models. Their strategy involves a multi-layered system of policies, technical setups, and established norms to prevent leaks and mitigate insider threats. This includes assigning strict confidentiality levels to information, using codenames for sensitive models, and implementing technical controls to restrict data access and sharing. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
RANK_REASON The article describes the internal policies and procedures of a research organization for handling confidential information, which falls under research and safety practices.
