Anthropic Limits AI Models Over National Security Order

AI safety firm Anthropic disabled its most powerful models for international users following a U.S. government directive ^[1]. The order aims to prevent the identification of critical software vulnerabilities by foreign actors ^[1][2]. This restriction specifically affects the Fable and Mythos model lines previously available to global developers ^[2].

The U.S. National Security directive reflects growing concern over the role of artificial intelligence in cyber warfare ^[1]. Officials believe high-end models could be used to automate the discovery of zero-day exploits ^[2]. While the move focuses on Anthropic, other AI companies face similar scrutiny as federal agencies probe potential harms and upcoming industry regulations ^[3].