AI-Powered Fraud in the Insurance Sector
It has been determined that millions of liras in unjustified compensation have been obtained in the insurance sector using fake damage photos generated by artificial intelligence.
Fraudsters in the insurance sector are filing claims by creating fake vehicle damage photos using generative AI tools that are indistinguishable from reality within seconds. Analyses have identified millions of liras in suspicious transactions.
Generating Fake Damage with Artificial Intelligence
Fraudsters use real photos of vehicles to create non-existent damage on them using artificial intelligence. This method specifically targets insurance processes that provide quick payouts for minor damages.
Visuals Difficult to Distinguish from Reality
Generative AI tools can create dents, broken headlights, bumper, and body damage within 10-15 seconds. Enhanced with light, shadow, and environmental details, these visuals are almost indistinguishable from reality when examined by the human eye.
Detection of Suspicious Files Worth Millions
Analyses conducted by FraudFlow, which provides technology support to insurance companies, identified 39 suspicious damage files utilizing AI-manipulated visuals. It was announced that the total amount of these files reached approximately 6.5 million liras.
A New Risk Area for the Sector
This situation creates a new risk area specifically in terms of damage review processes that rely solely on classic visual inspection. Industry representatives state that insurance fraud must now be addressed through new methods that incorporate artificial intelligence.
Using Technology Against Technology
Insurance companies are developing defense mechanisms against this threat. Thanks to technologies like ImageMap, visuals included in the damage processes are analyzed to detect manipulation traces that the human eye cannot spot.