Iran-linked actors used Anthropic's Claude to plan attacks on US naval forces

Anthropic has documented Iran-linked actors using its Claude AI for surveillance and potential attack planning against US naval forces, a confirmed instance of state-aligned actors leveraging commercial AI for military targeting. Additionally, Anthropic claims state-linked actors in China and Russia have used Claude for bioweapons research, and alleges its use in missile guidance in Yemen and global cyber operations. The extent of operational use and independent corroboration for all claims remains limited.