Aviva uncovered a record £230m of insurance fraud claims and uses AI tools to combat the growing problem.
The battlefield is changing and criminals are coming equipped with a new generation of tools. We are now in an environment where AI is being used not only to prevent fraud, but also to commit fraud.
The insurance industry has long dealt with opportunistic fraud. If you get into a car crash, you might suddenly need four new doors, or a minor slip could result in life-altering injuries. However, Aviva’s data shows that the nature of deception is becoming deeper, more sophisticated, and more elusive to the human eye.
Aviva deploys its own AI to uncover these elaborate conspiracies, lighting and putting out fires.
Combating insurance fraud factories with AI
Aviva reports that fraudsters are now using AI to create convincing fakes of road accident scenes. These are not clumsy jobs like Photoshop. These are detailed and plausible images that can easily fool insurance claims handlers who handle a large number of cases.
The same generative AI tools are being used to create fake documents, ranging from invoices for repairs that never happened to medical reports with no basis in fact. Scammers don’t need access to corrupt garages or networks of medical professionals to corroborate their stories. All you need is a subscription to an AI service and a little imagination. The AI takes care of the rest and creates an official-looking document that can pass a cursory inspection.
Individuals or small groups can now create evidence to support dozens of high-value claims without leaving their desks. How do we test reality when reality itself can be fabricated so easily and cheaply?
Aviva’s response was to build an AI-powered defense system that could operate at the same scale and speed as the threat. While the company is understandably tight-lipped about the exact architecture, we can summarize what such a system would need to do.
At its core, AI detectives perform pattern recognition at scale. AI sifts through millions of data points from current and past claims to learn what legitimate claims are and, more importantly, what legitimate claims are not.
As new claims come in, the system cross-references them all. Does the damage in the photo match the physical nature of the accident described? Do the timestamps on the documents make sense? Has this vehicle registration number appeared on other suspicious claims? Does the repair cost listed on the invoice match other similar repair costs in our database? This is a level of forensic analysis that would be impossible to perform manually on all the thousands of claims filed each day.
From organized crime to exaggerated claims
It is important to note that this is not just about organized crime groups. Part of this £230m figure is due to what the industry refers to as ‘claims inflation’.
Insurance claims are a more common type of fraud in which policyholders and service providers inflate bill amounts. For example, the garage may add unnecessary repairs to the estimate or overstate the value of items stolen in a burglary.
AI is proving to be a powerful tool here too. By analyzing a huge data set of repair costs and market prices, the system can instantly flag when estimated prices are outliers. You can compare the cost of replacement parts at one garage to the average for hundreds of other garages of the same make and model in the same area.
The goal of Aviva’s AI is not to completely refute a claim, but rather as an enrichment tool for human investigators. AI acts as a filter, sifting through the noise to reveal the most likely instances of fraud. This human-involved approach is essential to ensure fairness and prevent the system from becoming a black box making decisions without oversight.
What Aviva is doing provides a potential route for any customer-facing company in the age of generative AI. The same technologies that create these threats are also the most effective ways to combat them.
As everything from IDs to bills becomes easier to forge, the only viable defense is intelligent systems that can learn, adapt, and spot deception at a scale that humans alone cannot match.
See also: Weis Markets adds Instacart AI-powered shopping carts to stores
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expos in Amsterdam, California, and London. This comprehensive event is part of TechEx and co-located with other major technology events such as Cyber Security & Cloud Expo. Click here for more information.
AI News is brought to you by TechForge Media. Learn about other upcoming enterprise technology events and webinars.
