AI & Enterprise
Claude commands and Grok scouts in AI hacking that hit banks, report finds
Connecting AI models to rob banks is becoming increasingly realistic, the article says. An investigation found hacking group CyberCartel used multiple AI agents based on models from Anthropic, SpaceX and China’s Moonshot AI to automate attacks targeting about 400 financial institutions in Latin America. The Information cited cybersecurity startup Straiker as describing one of the first reports detailing how criminals automated intrusions. Agents allegedly stole passwords and data, assigned roles among models and bypassed safety controls.
