AI agents automate card skimming, 600K records stolen from 100+ e-commerce sites

A financially motivated threat actor has leveraged open-source AI agent frameworks to automate web skimming attacks against hundreds of online retailers, compromising over 100 sites and exfiltrating more than 600,000 credit card records. This marks a notable escalation in the use of AI to scale cybercrime, lowering the barrier for mass attacks. The full scope of affected merchants and the specific AI frameworks used remain partially undisclosed, but the incident signals a new threat vector for e-commerce security.