Car Insurance: L’Olivier Builds on AI Tech to Detect Fraud
FACTS
- L’Olivier Assurances plans to rely on Artificial Intelligence technologies and data science to fight car insurance fraud.
- These technologies will chiefly be used to automate and optimise suspicious damage claims.
- L’Olivier opted for a Software-as-a-Service AI solution by Shift Technology.
- This solution, called “Force”, bets on a platform for processing claims and support fraud-dedicated teams throughout the decision-making procedures.
- Following an initial cooperation phase, L’Olivier is making adjustments to implement these technologies in their detection process.
- By the end of the year, their teams will be able to automatically receive alerts warning them of sensitive claims, through an interface described as intuitive and user-friendly.
KEY FIGURES
- According to the Argus de l'assurance, fraud would have amounted to €2,5B in claims in 2014
- €219M were recovered by insurance groups
CHALLENGES
- Enhancing equity. Insurance fraud comes at a high cost for insurance groups, which triggers a knock-on effect on their prices and policies. Through fighting fraud, L’Olivier intends to pull prices downward, for all.
- Saving time. Shift Technology’s detection algorithms will help speed up claims processing activities.
- Betting on Machine Learning. One of Shift’s data scientist, fully assigned to L'Olivier, will keep enhancing Force’s scenarii and algorithms to further improve its accuracy.
MARKET PERSPECTIVE
- L’Olivier is only just getting started to work on these issues. The insurance group may take the concept one step further in the future and, for instance, integrate Shift Technology’s other solution called “Luke”. Luke automates claims management using AI to speed up processes, yet with more granularity and for a larger number of records.
- This isn’t the first time AI technologies are implemented for the sake of preventing fraud. Nice Actimize, for instance, recentlybuilt a Cloud-based AI-powered platform, especially meant to fight financial crime.