Anthropic Report: Claude AI Agents Weaponized for Cyber Extortion
Anthropic’s new Threat Intelligence report warns that agentic AI like Claude is being weaponized. It details “vibe-hacking” and a disrupted cybercrime ring that used Claude Code to extort data from at least 17 organizations in a month — including healthcare, emergency services, religious groups, and government — showing AI agents can amplify real-world harm.