Anthropic reports Yemen fighters used AI to seek guided weapons

AI firm Anthropic revealed that a cell of threat actors based in northern Yemen attempted to use its Claude chatbot to develop guidance software for rockets and missiles.

The cell tested a guided rocket that ultimately failed, after which they returned to the chatbot to troubleshoot the failure.

Anthropic stated that it banned the relevant accounts immediately upon detecting the unauthorized activity.