What happened
Anthropic says the actor automated reconnaissance, phishing infrastructure, malware rebuilds after detection, mailbox theft, hotel-WiFi DNS hijacking, and WhatsApp companion-device collection. Reported impacts include bulk export of drone-manufacturer mail and a North African government credential database with hundreds of thousands of national identity records. Humans directed targeting; Claude executed much of the operational work. Anthropic says it disrupted the observed activity and shared intelligence with partners. This is AI-enabled espionage, not an autonomous state-scale war.
