Chinese state-linked hackers tap Anthropic Claude for intrusion support, refine attacks with AI-assisted code work, and trigger major security responses across the U.S. and Europe.