IBM is aiming advanced analytics at fraudulent claims in the property and casualty (P&C) insurance industry. According to the National Insurance Crime Bureau, at least 10 percent of all insurance claims are either inflated or outright fraudulent. Consequently, the industry pays out as much as $30 billion in fraudulent claims each year.
Given the impact this fraud and abuse has on business and consumers, IBM designed its Loss Analysis and Warning Solution to help protect P&C insurance companies from illegal, unethical, or undesirable activities by policyholders, service providers and employees. The solution uses a combination of data mining technology, data visualization techniques and a powerful software engine.
Insurance companies can use the resulting information to detect unusual or suspicious behaviors such as inflated or unnecessary payouts, misuse by agents of policy discounts and unusual claims charges and billing practices from service providers such as repair shops, doctors and hospitals. The solution identifies normal behavior in these “peer” groups and makes comparisons to identify unusual behavior. For example it would flag an auto repair shop that was charging for towing on nearly all repairs, while other shops had towing charges less than 20 percent of the time.
IBM’s Advanced Analytics to Tackle Property & Casualty Insurance Claim Fraud (press release)