News

AI hacking tool breached five Korean banks, hit 25,000 at Shinhan alone

What's the deal? Investigators probing a wave of cyberattacks on Korea's major banks have found traces of ARTEX, an AI-powered autonomous penetration-testing tool. Shinhan BankDealroom has a profile for this one. Try Dealroom →, KB Kookmin BankDealroom has a profile for this one. Try Dealroom →, Hana BankDealroom has a profile for this one. Try Dealroom →, Yegaram Savings Bank and BNK Busan BankDealroom has a profile for this one. Try Dealroom → have all reported data breaches since last week.

How big is it? Shinhan Bank said about 25,000 customers were affected, with names, phone numbers, annual income and loan limits exposed. KB Kookmin reported 119 affected customers; Hana Bank said 89 had personal information exposed. Woori BankDealroom has a profile for this one. Try Dealroom → and NH NongHyup BankDealroom has a profile for this one. Try Dealroom → detected hacking attempts but confirmed no leaks.

What's new? ARTEX, a Chinese-language tool, automates tasks that normally require human expertise, such as scanning for vulnerabilities and building attack paths. It can connect to AI models from China's DeepSeek, as well as OpenAI and Anthropic, letting multiple AI agents analyse targets and intrusion routes.

Why now? The speed and scale point to automation. "It would be difficult for humans to carry out attacks this fast and on such a large scale across the financial sector in just a matter of days. It is highly likely that AI was used," said Lim Jong-in, a professor at Korea University's Graduate School of Information Security.

He compared the gap between human and AI-powered hackers to that between a runner and a Ferrari, adding that manual attacks of this size would take "weeks, if not months."

What could go wrong? Attribution is murky. Lim cautioned against blaming China simply because a Chinese-language tool was found. "The actor behind the attacks remains unknown," he said, noting that some Chinese AI models ship with weak safeguards, letting anyone download and modify them for malicious use.

The signal: AI can make hacking faster and broader by automating steps that once demanded significant human effort. Experts warn the breaches may mark the start of a wider shift in the cybersecurity landscape.

Read more: koreatimes.co.kr

Image credit: Visual Content

More top stories