AI Maps Out Over 3,400 Illegal Loan Shark Networks

[Anchor]

Illegal debt collection continues to plague borrowers, with lenders demanding thousands of times the principal in interest. By using AI to analyze these activities, investigators have uncovered deeply intertwined connections among illegal private loan shark networks.

Reporter Min Gyeongho has the story.

[Reporter]

In July, a man in his 40s, surnamed A, responded to a loan advertisement message that he normally would have deleted, borrowing 300,000 won.

He needed the money to cover living expenses.

[Illegal Debt Collection Victim: They said they would process it right away. Two or three weeks later, they told me to pay back about 480,000 won...]

However, when he failed to repay the loan on time, they slapped him with a late fee of 50,000 won per day, swelling the debt to 800,000 won.

With the late fees exceeding the principal amount,

Mr. A stopped transferring money to the lending agency after receiving advice from the Financial Supervisory Service that the practice was illegal and he no longer needed to pay.

That was when the threats and harassment began.

[Illegal Debt Collection Victim: They started contacting my wife and all my family members around me. They sent sexually explicit text messages. My son had no idea what those messages were about... He is a first grader, a first-grade elementary school student.]

Because the lenders use Telegram with servers based overseas, reporting them to the police has been of little use.

However, the Financial Supervisory Service has developed an investigation support system that uses artificial intelligence to analyze these syndicates.

The system treats entities as related organizations and creates a sort of map if information such as account numbers or mobile phone numbers used in crimes appears three or more times across different accounts.

[Kim Yu-hong / Senior Investigator, General Livelihood Protection Bureau, Financial Supervisory Service: When the exact same messenger and phone numbers are mentioned in separate victim reports, a line is drawn connecting them.]

This is a record of the Telegram account that contacted Mr. A, who suffered from debt collection harassment, as submitted to the Financial Supervisory Service.

Tracing these records reveals specific groups, and taking a closer look at these groups shows they were part of a larger organization.

As a result, approximately 3,400 networks have been identified.

This means that catching a single syndicate member will make it possible to track down the entire network and even other connected organizations, much like pulling up sweet potatoes by their vines.

The Financial Supervisory Service stated that once a law revision allows it to conduct direct investigations starting in January next year, it will actively utilize this system for investigations and share the technique with other investigative agencies such as the police.

(Camera: Lim U-sik | Video Editing: Jang Hyun-ki)
※ Please note: This article was translated by AI and may contain errors.