Scammers have begun actively using neural networks to fake information about traffic accidents in order to deceive insurance companies. This was stated by Arman Gasparyan, Head of the Loss Adjustment Unit at Rosgosstrakh. According to him, such offers can already be found on the Internet.
Gasparyan said that fraudsters offer services to create plausible images of allegedly damaged cars for just a few hundred rubles. They ask for a real photo of the car, guaranteeing to then send AI-generated photos of the car with damage.
At first glance, such services may seem like a joke or a reason to prank acquaintances, but, as Gasparyan emphasizes, if you study the ads in detail, you can see that it is proposed to take photos not only of cars, but also of other equipment, including trucks.
Insurers today are developing digital services and offering customers to inspect and photograph the damage themselves. Fraudsters are also trying to use this - which means that insurers need to analyze the photos attached to the application for payment even more carefully.
According to the expert, fraudsters also use AI to send pre-trial claims. Such letters often mention non-existent regulations and laws, which makes it difficult for lawyers to process them. Gasparyan emphasizes that such "hallucinations" of artificial intelligence can be misleading, forcing specialists to spend more time verifying data.
