A woman with two children, identified as A, was caught after forging documents to make it look as if her children had visited a hospital and receiving an insurance payout. The medical certificate and itemized receipt she submitted were reportedly so sophisticated that it was hard to determine whether they were forged. The insurer clawed back the payout made to A and canceled the policy.
A person in their 30s, identified as B, was also caught after forging medical records such as a medical certificate, an itemized receipt, and payment receipts to obtain an insurance payout. It was a case that might have been missed if the insurer had not noticed that the documents this hospital issued to other patients around the same time, as well as the hospital's seal and the doctor's signature, were different. The insurer filed a complaint against B, and police sent the case to prosecutors with a recommendation to indict.
According to the insurance industry on the 31st, insurance fraud using generative artificial intelligence (AI) has recently become a headache. As AI technology advances, forgery and alteration techniques have become so sophisticated that they are hard to distinguish from genuine documents.
A common tactic is forging medical records such as medical certificates and itemized receipts. People who did not actually receive treatment alter someone else's medical certificate and submit it, or inflate the treatment period written on inpatient or outpatient confirmation documents to claim insurance money. Overseas, there are even cases that go beyond simple document forgery to manipulating images. Some inflate photos of minor car scrapes into serious damage, or create images of non-existent diamonds to claim loss payouts.
As insurance fraud using AI increases, the overall scale of insurance fraud is also on the rise. According to the Financial Supervisory Service and the insurance industry, the amount detected last year reached 1.1571 trillion won, a record high. After first surpassing 1 trillion won in 2022, the amount rose to 1.1164 trillion won in 2023, 1.1502 trillion won in 2024, and 1.1571 trillion won in 2025, increasing every year. If insurers' revenue deteriorates due to fraud, premiums for ordinary policyholders will rise.
Insurers are drawing up countermeasures. KB Insurance analyzes the risk level of each claim and has built a system that automatically classifies claims with red flags. Kakao Pay Insurance is also using AI to analyze claim documents and repeat claim patterns. DB Insurance operates a big data-based insurance fraud investigation and analysis system to detect organized fraud.
There are also concerns that detection will become more difficult as AI technology becomes more sophisticated. An insurance industry official said, "Competition between forgery and detection techniques is inevitable," and added, "Ultimately, it is necessary to expand computer-linked systems such as 'Silsun24' so that hospital records are sent directly to insurers without consumers having to submit documents themselves."