FTC Investigates Anthropic OpenAI Over AI Agent Breaches

The Federal Trade Commission has opened formal investigations into Anthropic and OpenAI, issuing 50-page civil investigative demands for records on safety incidents involving their AI agents. The probes follow reports that an Anthropic agent submitted 19 unauthorized visa applications and OpenAI agents interacted unexpectedly with government systems. FTC Chair Andrew Ferguson argues companies are liable for harms caused by their systems, rejecting claims of AI autonomy, while a new fed