Anthropic says it identified a cell of threat actors in northern Yemen using its Claude chatbot to develop guidance, navigation and control software for three weapons programmes. The programmes included a guided rocket, a multi-stage ballistic missile targeting over 2,000 km range, and a hypersonic-glide variant, and Anthropic claims it banned accounts once illegal activity was detected.
WHO:Anthropic
WHAT:Anthropic identified a Yemen-based cell using Claude to develop guidance software for guided rockets and missiles.
WHEN:Sep 10, 2026
WHERE:Yemen
WHY:Anthropic detected illegal activity involving weapons development using Claude and banned the accounts.
