Hackers Use ARTEX AI Framework to Target Seven South Korean Banks, Exposing Thousands of Customers

Researchers report a campaign (late Sept–early Oct 2026) where attackers used an ARTEX AI-driven penetration testing platform to exfiltrate and expose data from multiple South Korean financial organizations; recovered artifacts include exposed Claude Code session histories, configuration files, proxies and scripts supporting reverse SSH, routing, and JNDI/LDAP-style exploitation, with estimated impact of ~66,000–67,000 individuals and ~2,200 companies and several IP-based indicators provided.