According to AI firm Anthropic, an Iran-linked cybercriminal group utilized the Claude AI assistant to draft attack recommendations against US naval forces. The revelation highlights growing concerns regarding the military application of advanced artificial intelligence.