From Enrollment to Compensation: AI Integrating into the Insurance Industry with Pre-Detection...

Samsung Fire & Marine Insurance Uses AI for Initial Claims Processing
Automatically Classifies Suspicious Fraud by Analyzing Relationship Maps
KB Insurance Utilizes AI in Underwriting
Expanded Coverage While Maintaining Loss Ratio

A system has been established to determine fraud at the early stage of compensation using artificial intelligence (AI) to prevent insurance fraud. AI is also being applied to automobile insurance underwriting to assess risk, creating an atmosphere where AI is permeating various parts of the insurance industry.

According to the industry on the 16th, AI technology is rapidly being introduced into the insurance industry, especially in non-life insurance companies. Samsung Fire & Marine Insurance's recently revamped insurance fraud prevention system, 'IFDS,' is a representative example. After building the first-stage system focused on post-detection in September 2021, the second-stage IFDS, centered on pre-detection prediction, has been developed and operated over seven months from November 2022 to last May.

The second-stage IFDS strengthened pre-detection of high-risk groups for insurance fraud based on scores for each suspect. The implementation of a relationship analysis system also improved the analysis of organized insurance fraud. Based on an index composed of various risk factors, the system calculates a risk score for suspected insurance fraud cases and provides detailed attributes and analysis results for cases with high scores.

In particular, by utilizing AI machine learning techniques, the risk level of customers suspected of insurance fraud is immediately provided on the work screen. It offers highly reliable detailed information on insurance fraud to detect cases suspected of multiple intentional accidents or organized insurance fraud at the early stage of compensation processing.

Additionally, a relationship analysis service linked to insurance fraud suspects is also provided. It is expected to be usefully employed to reveal collusion relationships among individuals and illegal companies related to insurance fraud. Suspected insurance fraud cases filtered through this analysis are automatically assigned to the Special Investigation Unit (SIU).

KB Insurance is actively using AI in the pre-subscription stage of automobile insurance. It was the first in the industry to introduce AI machine learning techniques into automobile insurance underwriting. Whereas underwriting work previously served as a barrier to filter out malicious cases, it now focuses more on identifying high-quality cases among those previously filtered out.

Since its first introduction in November last year, results have also emerged. KB Insurance's automobile insurance loss ratio was in the 76% range cumulatively from January to April this year. This is similar to the 75% range during the same period last year when movement was significantly reduced due to COVID-19. Despite expanding the market by accepting customers who would have been previously rejected, the loss ratio (the ratio of claims paid to premiums received) has been maintained stably.

Artificial Intelligence (Not directly related to the article) / Photo by Getty Images Bank