by Insurance fraud is any act committed to defraud an insurance process. It occurs when a claimant attempts to obtain some benefit or advantage they are not entitled to, or when an insurer knowingly denies some benefit that is due. The most common schemes include premium diversion, fee churning, asset diversion, and workers compensation fraud. Perpetrators in the schemes can be insurance company employees or claimants. False insurance claims are insurance claims filed with the fraudulent intention towards an insurance provider.
Insurance fraud is an illegal act on the part of either the buyer or seller of an insurance contract. Insurance fraud from the issuer includes selling policies from non-existent companies, failing to submit premiums, and churning policies to create more commissions. Buyer fraud, meanwhile, can consist of exaggerated claims, falsified medical history, post-dated policies, viatical fraud, faked death or kidnapping, and murder.
Insurance fraud has existed since the beginning of insurance as a commercial enterprise. Fraudulent claims account for a significant portion of all claims received by insurers, and cost billions of dollars annually. Types of insurance fraud are diverse and occur in all areas of insurance. Insurance crimes also range in severity, from slightly exaggerating claims to deliberately causing accidents or damage. Fraudulent activities affect the lives of innocent people, both directly through accidental or intentional injury or damage, and indirectly by the crimes leading to higher insurance premiums. Insurance fraud poses a significant problem, and governments and other organizations try to deter such activity.
Hard vs. soft fraud
Hard fraud occurs when someone deliberately plans or invents a loss, such as a collision, auto theft, or fire that is covered by their insurance policy in order to claim payment for damages. Criminal rings are sometimes involved in hard fraud schemes that can steal millions of dollars.
Soft fraud, which is far more common than hard fraud, is sometimes also referred to as opportunistic fraud. This type of fraud consists of policyholders exaggerating otherwise legitimate claims. For example, when involved in an automotive collision an insured person might claim more damage than actually occurred. Soft fraud can also occur when, while obtaining a new health insurance policy, an individual misreports previous or existing conditions to obtain a lower premium on the insurance policy.
Types of Insurance Fraud Schemes
Sellers
- Premium diversion: An example of premium diversion is when a business or individual sells insurance without a license and then does not pay claims.
- Fee churning: When intermediaries such as reinsurers are involved. Each takes a commission that dilutes the initial premium so that there is no longer any money left to pay for claims.
- Asset diversion: The theft of insurance company assets, such as, for example, using borrowed funds to buy an insurance company and then using the acquired company’s assets to pay off the debt.
Buyers
Attempts to illegally reap funds from insurance policies by buyers can take on a variety of forms and methods. Insurance fraud with automobiles, for instance, may include disposing of a vehicle and then claiming it was stolen in order to receive a settlement payment or a replacement vehicle.
The original vehicle could be secretly sold to a third party, abandoned in a remote location, intentionally destroyed by fire, or pushed into a river or lake. If the owner sells the vehicle, they would seek to profit by pocketing the cash, and then claim the vehicle was stolen in order to receive further compensation.
Fraud prevention
- Implement a foundational framework
A foundational framework should reflect a fraud-detection strategy that addresses such questions as: How can we check all claims for fraud but ensure fast claim processing? How can we identify fraud before a claim is paid? How can we improve fraud investigation efficiency? How can we keep track of changing fraud behaviours? How can we reduce false positive signals? And finally: What is the best approach to automate the fraud-detection process and predict the likelihood of fraud? Implementing a foundational framework enables management to make better decisions about priorities, resource deployment and investments.
A foundational framework can range from an “out-of-the-box” solution that automates the institutional knowledge of your claims professionals and enables workflow management to full social networking analysis of the parties involved in a claim. From there, insurers can add a multitude of scoring engines, third-party data captures, criminal history lookups and many other tools. An important aspect of fraud detection is having a culture in your claims staff that emphasizes the importance of recognizing, identifying and investigating suspicious claims. Empower your staff to be involved, and then the tools you deploy will function much more effectively.
- Know the relative level of fraud potential
Knowing the relative level of fraud potential for every type of claim allows the best, and quickest, action to be taken to maximize special investigative unit (SIU) efficiency and savings. With limited resources to devote to fraud, it is important to make sure your investigations can be focused on the items that have the greatest potential for cost avoidance and successful identifications. For example, a theft claim involving the suspicious disappearance of expensive jewellery has a higher potential for being fraudulent than a stolen smartphone or laptop. Examples of common false claim schemes include deliberately destroying property and misreporting the cost of auto repairs.
- Use data analytics to detect fraud
Fraud comes in all shapes and sizes. In general, insurance fraud can be divided into two categories: criminal fraud, which is perpetrated by professionals habitually trying to milk the system; and cultural fraud, which is a genuine claimant being opportunistic or exaggerating a claim.
Data analytics can be applied to detect fraud. By analyzing past fraud, insurers can use predictive modeling to produce what is called a “Suspicion Score,” a value for the propensity of fraud. The process works like this: Adjusters simply enter data, and claims are automatically given a Suspicion Score to indicate the likelihood that fraud has occurred. The technology behind this involves utilizing data-mining tools and applying quantitative analysis.
Even with automation and data analytics, the weakest link in fighting fraud can be your own employees. The importance of checks and balances cannot be stressed enough.
- Continually review and rescore claims
Success in combating insurance fraud comes from persistence and good timing. Above all, apply your arsenal of tools including data analytics and predictive modeling early and often. Claims should be continuously monitored for fraud potential. As an insurance company, it is imperative that you target the right claims, at the right time, with the right tools. Luckily, predictive modeling and advanced analytics are coming into play as essential tools for fighting insurance fraud. These tools can be automated, preventing the need for hands-on manual analysis.
By continuously reviewing and rescoring claims using Suspicion Scores, insurers can detect patterns that reveal fraud. Some claims score high immediately at first notice of loss, prompting your SIU to get involved immediately. For others, high scores do not show up until after the claim has been collected.
Monitoring Suspicion Scores has been shown to be more accurate and more effective than traditional fraud-detection methods. But again, the key is to not rely solely on technology to do all of the heavy lifting human analysts are required to initiate action after the suspected fraud has been flagged, and your people must follow through with appropriate measures. This is where training employees to identify fraud becomes an important piece of the overall fraud-detection puzzle.
- Adopt a layered approach
In the world of IT, a “layered approach” refers to using a variety of tools and technologies to tackle a challenge. In detecting insurance fraud, this means throwing the kitchen sink at the criminals, but doing it in an organized, well-considered fashion.
Fraud is a complex, multifaceted problem, and no single method can detect all fraud. Each fraud-detection method needs to be crafted to address a specific area. Different rules and indicators are needed for different types of policies and claims. Plus, fraudsters hide in multiple databases, so fraud-detection methods must search them all. Because of the complexity of fighting fraud, it is advisable to bring in outside expertise to help formulate a framework and implement the technology, tools and methods needed to deal effectively with fraud.
The modern insurance organization has a number of technology tools at its disposal to detect fraud. For example, videos, photos and even livestreaming can be used to document evidence at a car crash or crime scene. It’s difficult for the average person to fake a video, especially when the device’s location access is turned on. A virtual gold mine lies within unstructured data, and it is imperative to collect, organize, index and mine the data to detect fraud. Always remember: You can’t claim what you can’t prove.
- Revise based on market conditions
Criminals are ever resourceful, so always be ready to quickly adapt to changes in the ways fraud is undertaken, as well as changes in your industry. For example, professional criminals are sophisticated enough to become familiar with the analytical approaches that insurance companies use to detect fraud, and to change their tactics when committing fraud. As fighting fraud becomes more proactive, insurers must spot new fraud trends early and take steps to stay ahead of the bad guys.
Your everyday policyholders may also try to be more creative with their insurance claims when the economy is in a down cycle. Keep your claims staff aware of the type of market conditions the policyholders are facing so the staff can be on the lookout for new and inventive fraud attempts that may be unknown to the software in place.
