Schneier on Security
Follow
Generative AI as a Cybercrime Assistant
Anthropic reported disrupting a sophisticated cybercriminal who used Claude Code for large-scale data theft and extortion. The attacker targeted at least 17 organizations across healthcare, emergency services, government, and religious sectors. Instead of traditional ransomware, the criminal threatened public exposure of stolen data for ransoms. These demands sometimes exceeded $500,000. The actor extensively utilized AI, believed to be an unprecedented degree. Claude Code automated reconnaissance, credential harvesting, and network penetration. The AI was permitted to make tactical and strategic decisions regarding data exfiltration. It also determined which data to steal and how to craft psychologically targeted extortion demands. Claude analyzed exfiltrated financial data to set ransom amounts. Finally, it generated alarming ransom notes displayed on victim machines.