Eighty percent of insurers use predictive modeling to fight fraud, study

The use of predictive analytics by insurers to fight fraud has reached an all-time high with 80% of insurers using predictive modeling to detect fraud, up from 55% in 2018, according to a study from the Coalition Against Insurance Fraud and SAS.

The State of Insurance Fraud Technology study surveys the industry’s fraud tech professionals and includes research that examines the use of anti-fraud technologies, related challenges and anticipated technology investments. Results are based on a 20-question survey sent to 100 Coalition members in October 2021

David Hartley, director of insurance solutions at SAS said in a statement: “The shifts we’ve seen since the 2018 study emphasize the increasingly sophisticated technologies needed to foil insurance fraudsters’ criminal exploits. Predictive modeling is up 25%. Text mining has nearly doubled, jumping from 33% to 65% in three years. These findings prove that, even as COVID has fueled rampant fraud, insurers are agilely stretching their advanced analytics and AI capabilities to counter rapidly changing threats.”

Research also shows rapid adoption of predictive algorithms and digital identity technology will stem the pandemic-driven fraud deluge. Insurance fraud causes more than $80 billion in losses annually in the U.S. alone, according to the report.

Anti-fraud technology is evolving rapidly with the use of artificial intelligence, geotargeting, automation and other advancements in information technology to fight fraud, according to the report. The study showed that limited IT resources (68%), data integration and & poor data quality remain the most significant implementation challenges in 2021.

“We know that criminals are using advanced technology at scale to steal personal information and plunder billions of dollars from insurance companies each year,” said Kim Kuster, principal business consultant in SAS’ Global Security Intelligence Practice, in a statement. “Wider adoption of emerging technologies and deeper investment in human- and machine-powered fraud fighting capabilities will help turn the tide of fraud flooding the domestic and international insurance markets.”

Additional findings in the study include the following:

• Insurers are diversifying their data sources. Beyond relying on their internal data, insurers are turning to industry fraud-watch lists (88%), public records (79%), third-party data aggregators (55%), social media data (48%) and data from personal devices (15%). The use of unstructured data soared from just under half in 2018 to 81% in 2021.
• Insurers are using photo analysis technology, up from 49% in 2018 to 81% in 2021, to authenticate claim damage, identify digitally altered images and index pictures submitted in other claims. 
• New anti-fraud technology is creating efficiencies in investigative processes, but the resources insurers are dedicating to internal and external investigative teams are insufficient to keep pace with the billions in fraud committed each year. Limited IT resources were the top anti-fraud challenge, cited by 68% of respondents.
• In a category entirely new to the 2021 survey, 40% of respondents cited the use of identity verification software.
• The report highlights the growing use of technologies like physical and behavioral biometrics, computer vision analysis, robotics, blockchain and virtual and augmented reality in the insurance sector.