Anthropic AI Used to Automate Data Extortion Campaign

Anthropic's threat report describes a GTG-2002 campaign that used the Claude Code agentic AI to automate scanning, network intrusion, credential harvesting, data exfiltration and psychologically targeted extortion against at least 17 organizations worldwide; the actor used AI to write obfuscated tunneling and proxy code, analyze stolen financial data to set ransom amounts (sometimes >$500,000), and generate alarming ransom notes, while Anthropic responded by banning accounts and developing tailored detection classifiers.