Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

IncidentAI-enabled attacks

Seoul police arrest a former penetration tester whose ring used ARTEX to steal 8.2 million records from more than 90 organizations

HighSeoul Metropolitan Police Agency

The Seoul Metropolitan Police identified seven suspects behind breaches of more than 90 organizations in South Korea, among them educational institutes, hospitals, large corporations and golf courses, from which about 8.2 million records were stolen. The attacks began in September 2025 with conventional tools; in late August 2026 the group adopted AI utilities to expand them, with the ARTEX penetration testing suite, which police describe as an autonomous framework, as the main tool. BleepingComputer notes it is the same framework used against several of the country's banks. In 38 incidents the group demanded 360 million won, about 270,000 dollars, in return for not leaking the data. The 32-year-old leader, described as self-taught but reported by local media to have worked at a cybersecurity company from 2019 to 2022, was arrested on October 15, formally detained on October 17 and referred to prosecutors on October 23; the other six are under investigation and not detained. A victim's complaint in June 2026 started the case, and investigators traced the suspects through the breached organizations' server logs, then used an in-house AI tool to analyze the evidence. Police call it the first arrest in Korea for large-scale hacking and personal data theft that exploited generative AI, and say that people who are not professional hacker organizations can now hack on a large scale with AI's help. They are checking whether any of the data has leaked on the dark web.

Replay the attack step by step in the Learning Center

How it happened

  1. A small group with one trained penetration tester ran conventional intrusions for a year, then added AI tooling to scale up.
  2. The ARTEX framework automated the intrusion steps against organizations that had no reason to expect a targeted attack.
  3. Stolen records became extortion demands: pay, or the data leaks.

Impact

About 8.2 million records taken from more than 90 organizations, and ransom demands totalling 360 million won across 38 incidents.

The fix

Nothing to patch on the victims' side beyond the usual: patched internet-facing systems, phishing-resistant MFA on remote access, and logs kept long enough for investigators to use. The arrests show that server logs still identify attackers who use AI.

Lessons

  • The same autonomous attack framework that hit the country's banks served a seven-person extortion ring against golf courses; capability has stopped being the barrier.
  • Schools, hospitals and clubs hold records worth extorting, and automated tooling makes them cheap to reach.
  • Server logs on the victims' side were the trail; keep them, and keep them long enough.

Sources