Safety in AI

Anthropic Rolls Out Claude Gov Models for Top-Level U.S. Security Agencies

Looser operational constraints let the models engage with classified inputs for intelligence analysis in top-level security environments