Zscaler tracks abuse of ARTEX AI pentesting framework against South Korean banks

Zscaler ThreatLabz has detailed a campaign in which a financially motivated, Chinese-speaking threat actor abused ARTEX, an open source agentic penetration testing framework, to target South Korean financial institutions. The activity was first observed in late September 2026 and led to the theft of personally identifiable information belonging to roughly 67,000 people.
The actor used ARTEX, which orchestrates large language models to automate reconnaissance and exploitation, against peripheral internet-facing systems including customer inquiry portals and employee mobile application backends. Zscaler said the stolen records included names, phone numbers, income figures and loan limits, and that the actor sought advice on selling the data through Korean-language Telegram channels.