Iran-Linked Actor Uses AI to Target U.S. Navy

• •Iran-linked actor used Anthropic's Claude AI for targeting U.S. Navy forces.
• •Exploited publicly available data to compile targeting handbooks and research vulnerabilities.
• •Anthropic disrupted the operation and banned the associated user.

A threat actor linked to Iran exploited Anthropic's AI model, Claude, to gather and analyze public information for targeting U.S. Navy forces in the Middle East. The operation involved compiling a roster of U.S. personnel from military photographs, collecting ship and aircraft transponder data, and generating scripts for commercial satellite imagery queries. The actor also researched vulnerabilities in shipboard systems, including flaws in maritime satellite communications and Cisco equipment. Anthropic discovered and disrupted this activity, banning the associated user and developing new detection methods. The report detailed incidents from December 2025 to August 2026, highlighting the misuse of AI in various cyber operations. The U.S. Navy has since warned personnel to enhance security measures against potential threats. Anthropic's report also noted similar misuse by other state-linked actors, indicating a broader trend of AI exploitation in cyber operations.