Insurance fraud is big business — and it's impacting everyone. Fraud costs U.S. insurers an estimated $308.6 billion annually, according to a report by the Coalition Against Insurance Fraud (CAIF) — up from $80 billion when the CAIF first conducted the study in 1995. Invariably, these get passed on to the consumer, further hurting an insurer's competitiveness.
As the misuse of AI technology leads to more deceptive fraud schemes like deepfakes—using AI to manipulate, generate, or swap faces and recreate voices—institutional knowledge is dwindling from an aging and shrinking claims-handling workforce. Couple this with the high volume of claims, and many adjusters and analysts have difficulty scrutinizing each claim for fraudulent activity. Additionally, there's less emphasis on tracking cost savings related to fraud investigations. This further discourages taking aggressive action against suspect claims.
According to Kevin Lederer, Senior Vice President of the Special Investigation Unit (SIU) at Davies, to successfully combat the new era of fraudsters, it requires a multi-prong approach (which, yes, incorporates AI and machine learning) that could have cost benefits beyond avoiding bad claims.
A Multi-Pronged Approach to Fighting Fraud
The industry is first fighting back with a multi-pronged approach by incorporating AI-powered fraud detection software into claims systems to identify potential fraud throughout the claims cycle. Through machine learning, this technology identifies suspicious patterns and irregularities in massive amounts of data at speeds that exceed human capabilities. Lederer says that AI-powered machine learning brings a new level of accuracy to fraud detection, enabling the identification of novel and evolving fraud schemes that might otherwise go unnoticed.
"It all adds up to greater cost savings by reducing unnecessary or fraudulent claim payments," says Lederer.
But the cost savings don't end there. Automation of manual tasks through AI leads to additional cost reductions due to increased efficiency, not to mention freeing up time so claims adjusters can concentrate on fraud detection and defense strategies. AI and machine learning also enhance data analytics by providing richer and more diverse datasets, opening up new opportunities for faster and more effective fraud defenses and claim closures. For instance, AI and machine learning can run reports and identify suspicious patterns and irregularities, partially automating a process that could otherwise be time-consuming for the time-strapped workforce.
Fraud Prevention Beyond AI
Besides technology, insurers must also fight back by training claims staff beyond the minimum required for regulatory compliance. Regular and comprehensive fraud training helps claims staff become more aware of current fraud trends and empowers them to take proactive steps to combat it. Also, recognizing that victory against fraud requires a collaborative effort, insurers should partner with the National Insurance Crime Bureau and the International Association of Special Investigation Units.
Still, Lederer says there's more to be done because although the insurance fraud-defense industry remains active and constantly evolving, so does the fraud. While AI machine learning is somewhat limited in many ways, hopes are high that it will soon be able to expand its investigative use to spot patterns on social media, conduct background checks, conduct medical histories, and even conduct surveillance.
"When it comes to battling fraudulent claims, the best defense is a good offense," says Lederer. "Insurers that incorporate comprehensive training with strong industry partnerships and, above all, strong AI-powered technology will root out fraud and put a cap on financial losses."
For more information on Davies fraud detection, click here.
© 2024 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.
