‘Unprecedented’ Cybercrime Surge Linked to AI Automation, Says Anthropic

Anthropic reported that a hacker exploited its Claude chatbot to automate what it called an “unprecedented” AI-driven cybercrime spree, targeting at least 17 companies over three months. The hacker used Claude Code to identify vulnerabilities, create malware, analyze stolen data, and even draft ransom notes, demanding $75,000–$500,000 in bitcoin. Victims included a defense contractor, a financial institution, and healthcare providers, with stolen data ranging from Social Security numbers to sensitive defense files. Anthropic declined to name the companies but confirmed additional safeguards were introduced. The case highlights rising risks as AI lowers barriers for sophisticated cybercrime operations.