AND-E Working With Partners on Fraud Detection, Utilising AI

Car insurance is gradually evolving into something more personalised, usage-based and data-driven. But there’s a long way to go. One aspect of the quote process is identifying “fronting,” fake jobs, fake ID in general, or spotting claims histories associated with named drivers. Here’s some news from AND-E as they streamline claims data workflow;

AND-E (Aioi Nissay Dowa Europe) has worked in partnership with the Aioi R&D Lab- Oxford and its partner, Mind Foundry, to develop a powerful AI solution designed to combat the ever-evolving threat of insurance fraud. Aligned to AND-E’s purpose to make tomorrow´s journeys simpler, safer and smarter, the tailor-made fraud investigation solution, utilising leading-edge AI-driven techniques, enables quicker claims handing, reduces claims costs and has improved overall fraud detection by 120%.

Since deployment in 2022 the new AND-E AI fraud investigation solution increased automated referrals retained by the AND-E fraud department by 500% and has maintained that performance reliably without the need for any manual retraining in 2 years of operation. Working on fewer claims, the team can close cases more quickly, increasing efficiency, reducing the cost of claims and ultimately reducing customers’ premiums.

“Fraud is a constant and ever-evolving threat in the global insurance industry,” commented Jan Martin, Head of Third Party Claims at AND-E UK. “We recognise the importance of staying ahead of these challenges to protect our customers and our business.

“Our previous traditional rules-based fraud detection model would flag and triage potential fraudulent claims, however as the fraudsters got more sophisticated and techniques evolved, there needed to be a new approach. We worked closely with Mind Foundry to develop an explainable AI solution that fit our unique business requirements and most importantly, would gradually get better and better at learning fraud detection patterns.

“Utilised correctly, AI is invaluable, however, it doesn’t replace the expertise of claims professionals. The solution we have built with Mind Foundry allows collaboration between claims handlers and AI, increasing efficiency, cutting costs and providing customers with an improved and seamless journey.”

At the heart of the fraud investigation solution is a powerful, continuously learning predictive model trained on a vast dataset, including over 20 million unstructured documents, handwritten notes, and historical claim data. The model looks at all incoming claims and, at any given point in time, can assign each claim a fraud score based on a set of features jointly developed by AND-E UK’s fraud specialists and Mind Foundry’s data scientists and engineers.

Crucially, the model also learns continuously. This lends a critical advantage to counter-fraud efforts because the behaviour of people trying to commit fraud is ever-changing and ever-adapting to any countermeasures. A successful model is one that stays ahead of those changes, or at least it adapts to them quickly.

In the past year AND-E has been able to reduce the overall claims cost to the business by recording a 2% saving on its capped Third-Party indemnity spend. And the ever-improving nature of the new solution means the results are expected to improve further year-on-year.