Open-Source Chinese AI Agent Breached Seven South Korean Banks, Exposing 65,000 Customer Records

Cybersecurity researchers have confirmed that hackers used ARTEX AI, an open-source autonomous penetration-testing system built on large language models, to orchestrate coordinated attacks against seven major South Korean financial institutions in late September and early October 2026, exposing the personal data of over 65,000 customers. This marks the first confirmed large-scale deployment of an AI agent against critical financial infrastructure, signaling a qualitative shift in the threat landscape.