AIOI R&D LAB-OXFORD developed an AI system to detect fraudulent claims in auto repair costs. Aioi Nissay Dowa Insurance will use this system to understand trends in repair cost claims from various auto repair shops, enhancing its damage assessment capabilities.

Background

Combatting fraudulent insurance claims is a critical issue for the entire damage insurance industry, prioritizing customer protection and the healthy, stable operation of the insurance system. Recent revelations of improper auto insurance claims by major used car sales companies have made it imperative for insurance companies to enhance their damage investigation structures. In response, Aioi Nissay Dowa Insurance, in collaboration with Mind Foundry, an AI venture of Oxford University, and their joint research institute Aioi R&D Lab-Oxford, has developed this system to strengthen investigative procedures by understanding the trends in repair cost claims submitted by repair shops.

Overview

About the System

1. The system visualizes the billing trends of each repair shop, differentiating between those without suspected fraud and others. It analyzes each shop's alignment with the 'Fraud Suspicion Model', created from previously identified fraudulent estimates.

2. It scores the likelihood of fraudulent claims, displaying higher scores for shops with such tendencies. The basis for score calculation is also shown, clarifying key investigation points for each repair shop.

Application of the System

For repair shops with high scores, the system will facilitate re-examination of past agreements, strengthening of inspection standards, and implementation of targeted investigations and evidence preservation requests before repairs.

Future Developments

The system will continuously learn from the latest repair and fraudulent claim data to adapt to changes in vehicle characteristics and repair methods. Ongoing algorithm refinement will be considered as part of the efforts to eradicate fraudulent claims.