business trends

Fraud Is Not A Cost of Doing Business

And emerging-tech is here to prove it

by Karen Pauli

Ms. Pauli is a principal with Strategy Meets Action (SMA), a leading strategic advisory services firm exclusively serving the insurance industry. Reprinted with permission. Visit here.

March 27, 2017 — Fraud has long been a significant problem for the insurance industry – actually since the very beginning of insurance at Lloyd’s coffee house.

The Coalition Against Insurance Fraud indicates that 5-10% of claims costs are related to fraud, with over 30% of insurers reporting as much as 20% of claims costs being related to fraud.

Fraud is a lucrative business for fraudsters, and perpetrating fraud becomes more creative every day. I am pretty certain that most heads of Special Investigations Units (SIU) feel that “fraudster” should be a job category within the Department of Labor … the focus on committing fraud is so relentless by some, it is almost a profession!

In my insurer career, I was a technical advisor to an SIU. I have always felt that was probably the best job assignment I ever had. The investigators were all ex-law enforcement – big city police officers and State Troopers with some FBI agents thrown in for good measure. They told the best stories about chasing down bad guys! Underlying it all, however, was frustration. Detecting fraud is hard. Finding the fraudsters and prosecuting them, even harder. Current estimates are that only 1.5% of cases are prosecuted. Unfortunately, some insurers have an attitude about fraud that borders on: “It’s just a cost of doing business.” However, that attitude cannot persist in today’s business environment where every dollar of claims costs must be acutely managed to maximize very thin bottom-line margins.

Detection, Mitigation & Prevention

The recent SMA research brief, Fighting Fraud with Advanced Technology: Detection, Mitigation, and Prevention, recounts the historical and current path of fraud detection, starting with the “gut feel” of seasoned claims adjustors. Then, along came business rules which allowed for uniformity and some automation. Today, predictive analytics and link analysis are the leading solutions for fraud detection. In particular, link analysis is an effective way to find fraud rings that attempt to hide within large claims volumes using technology to change their personas.

Big data and emerging technologies such as artificial intelligence (AI), behavior science, and behavioral analytics hold the promise of allowing insurers to get out in front of fraud.

Ironically, the new reality for insurers is that the more digital they become, the easier it is for fraudsters to hide and reinvent themselves. Fully automated, online new business applications allow fraudsters to gain access to coverage. Electronic claims submissions permit individuals, including unscrupulous doctors and lawyers, to submit “documentation” that payments are warranted. No insurer is going to stop their digital initiatives because of this. However, insurers need to augment business rules, predictive analytics, and link analysis with emerging technologies in the fight against fraud.

Telematics can assist adjusters, for example, in determining if a vehicle in question was in the location alleged at the time of the loss, or if the reported injuries actually equate to the crash details or appear to be fabricated. Telematics aren’t just for rating! Wearables can do the same thing relative to individual workers. Could a severe injury claimed from a fall actually have occurred given the dynamics of the fall?

Big Data and AI

Big data and emerging technologies such as artificial intelligence (AI), behavior science, and behavioral analytics hold the promise of allowing insurers to get out in front of fraud. The clear problem that SIU investigators have, even with link analysis and predictive analytics, and certainly with business rules, is that they are always chasing the fraudsters after they have gotten claim payments. It is true that predictive analytics and link analysis can minimize the number of fraudulent payments the fraudster obtains, but, the fact is that the bad guys get themselves into the payment queue and then the alerts and flags go up.

Big data, AI, and behavior analytics have the great potential to cut off the fraudsters before they get a claim payment. And, we don’t know what we don’t know when it comes to AI and behavioral analysis – whole new worlds of fraud fighting capabilities may arise out of new insights.

I would dearly love to reconnect with the SIU team I worked with “back in the day.” It would be amazing for them to see what current predictive analytics and link analysis in an automated fashion can do, where they once applied sweat and elbow grease to accomplish whatever they could with precious few positive results … and to brainstorm outcomes aided by telematics, wearables, AI, and behavior analysis. The most amazing thing for them to witness is that current and future fraud-related technology investments combined with the honed skills of SIU investigators can generate significant ROI and change the attitudes about fraud being another cost of doing business!