Fraudulent insurance claims are growing in volume, variety and velocity. According to the Insurance Information Institute, the estimated cost of property and casualty insurance claims fraud is $32 billion in the U.S. alone. These losses are only growing as the world economy is gripped by an alarming health crisis and a severe recession. Legacy fraud detection systems that apply simple fraud detection rules on aging set of known fraud types and scenarios and siloed data limit insurers’ ability to reduce fraud and uphold the experience of legitimate policyholders. Detected fraud rates are usually below 1% across lines of business while false positive rates are above 40%. Increasing volume of false positives and cumbersome manual processes reduce triage and investigative efficiencies.
The IBM Financial Crimes Insight for Claims Fraud solution is designed to help reduce losses from Property & Casualty (Auto / Motor, Home) claims, Medical Provider claims, etc., by applying cognitive analytics to identify complex and changing fraud patterns. The solution ingests first party claims data and 3rd party entity data and infuses AI into a pipeline of advanced analytics such as entity resolution, graph analytics, text analytics etc. formulated by industry experts. The solution uses a combination of multiple analytic techniques including Machine Learning models and deterministic rules to generate an overall risk score for each claim processed, improving the speed and accuracy of claims adjustment and investigation.
Join IBM and a select group of your industry peers as we discuss best in breed data management, AI and governance tools that will take fraudulent claims detection to another level.