Seoul police have arrested seven people accused of using generative artificial intelligence to break into about 90 company systems and steal more than 8.2 million pieces of personal and internal company data, in what investigators describe as South Korea's first large-scale AI hacking case. The Seoul Metropolitan Police Agency announced the arrests on October 26, after a 13-month investigation covering September 2025 to October 2026.
The 32-year-old ringleader was referred to prosecutors on October 23 while in detention, police said. The other six suspects are being investigated without detention. All seven face suspicion of violating South Korea's information and communications network law, along with extortion and attempted extortion.
How the scheme worked
According to police, the ringleader did not major in information technology and taught himself hacking techniques through online blogs and other material. He had previously worked at a security company, where he was responsible for mock hacking exercises designed to test defences.
Investigators say he began by directly operating automated tools that probed websites for vulnerabilities, then expanded his reach considerably once he started using generative AI. Police also confirmed the use of an AI autonomous penetration-testing system, a tool they say has also appeared in recent personal-data leaks affecting the financial sector.
Yonhap News Agency, citing police, reported that AI tools were used to hack 60 of the targets. The victims were not limited to large corporations: they included private academies, hospitals, golf courses, funeral service providers and universities, as well as affiliates of bigger companies.
Extortion demands and the scale of losses
Police say the group threatened 38 companies and demanded a combined 360 million won, about $272,000, warning that stolen data would be leaked if they did not pay. Threatening emails reportedly warned victims that refusal could damage their corporate image and expose them to fines or lawsuits from customers.
The custody referral for the ringleader covers about 900,000 pieces of personal information allegedly stolen from 15 companies that reported incidents to police. Investigators say they are still examining allegations involving 75 other affected companies that have not filed reports, and are looking into whether more companies were threatened and larger sums demanded.
Why police call it a first, and a warning
Police described the case as the first arrest in South Korea connected to large-scale hacking and personal-information theft using generative AI. In its announcement, the Seoul Metropolitan Police Agency said individuals without the backing of a professional hacking organisation had been able to attack large companies with AI assistance, adding that AI is lowering the barrier to entry into cybercrime.
Bae Sang-kyun, head of the Korea Institute for Criminal Justice Policy, said AI-driven hacking follows a different route from standard human-led attacks and can expose vulnerabilities across many systems at once. Authorities have recently been on heightened alert over cybercrime suspected of using AI tools, including a series of high-profile attacks on local financial institutions.
For companies holding large volumes of customer information, police say the case underlines a shift already visible across the region: the tools needed to probe, breach and monetise a system are no longer confined to highly specialised groups, and defences built for conventional attacks may miss the new routes AI-assisted intruders take.



