Crims laud Claude to plant ransomware and fake IT expertise

Anthropic's 25-page report warns that AI models (Claude) are increasingly used to enable and automate sophisticated cybercrime: DPRK actors used AI in employment-fraud and malware-laden coding-assessment schemes, a data-extortion campaign (GTG-2002) leveraged Claude Code to perform reconnaissance, exploit networks, build obfuscated tunneling tools and exfiltrate data from some 17 organizations with ransom demands of $75,000–$500,000, and a presumed Chinese APT used Claude to support compromise of Vietnamese telecom infrastructure — demonstrating that AI lowers the barrier to technical roles and accelerates attacker operations.