High-Tech Fraud Fighting

Q: How big a problem is fraud, in terms of money, resources and efficiencies? What are some of its sources and causes?

A: Gary Gummig, NAIC: The scope of the problem is huge. Fraud is about a $96 billion-a-year problem, and much of this is from false claims. This figure mostly represents consumer-to-company fraud. From a regulatory standpoint, the primary impediment preventing regulators from detecting fraud is a lack of data sharing among functional regulators in the financial services industry. Companies themselves have pretty sophisticated data mining tools to allow them to look for trends of fraudulent activities, but a key to detecting fraud is for insurers to be able to share information with each other and with other financial services industry participants.

A: William D. Barry, Docucorp International: According to the Coalition Against Insurance Fraud (CAIF), insurance fraud is a growing problem across the country, especially in areas like New York and southern California, where up to 50 percent of all auto insurance claims are fraudulent. The CAIF estimates insurance fraud totals $80 billion annually. The problem is so severe that most states have passed requirements that insurance companies establish Special Investigative Units (SIUs) to combat fraud because it increases costs, which results in increased premiums to the consumer.

A: Thomas J. Mulvey, The Prudential Insurance Company of America: The National Insurance Crime Bureau estimates that $30 billion in claim payments and related misrepresentations is the annual ante for insurance fraud. That is a large sum of money and therein lies the attraction for those with larceny in their hearts. There is, however, some good news.

First and foremost, the vast majority of policyholders are totally honest. Secondarily, we are experiencing a greater public awareness of insurance fraud and part of that enlightenment is the understanding that the consumer is shoulder-to-shoulder with the insurance company as a victim. As a result, pointed criminal investigative efforts are in place at the federal, state and county level.

Q: What are the key processes to be addressed, and how should they be addressed, from a technology standpoint?

A: Gummig: For years, the NAIC has been trying to get access to the FBI criminal database. Banks and the securities industry have access to this database, but the insurance industry does not. The Anti-Fraud Network bill has been introduced (HR 1408), and is designed to give insurance regulators access to this information, which will help immensely in preventing fraud. This would provide a network for shared information among insurance, securities and banking industries. With the convergence of the financial services industry, it becomes increasingly difficult to draw a line when it comes to selling and regulating services offered by financial institutions, securities firms, and insurance companies. The Internet could be the perfect tool to reach across the networks and into the databases to facilitate the sharing of pertinent information.

A: Barry: Claims provide the greatest source of information on fraud and are the starting place for most fraud detection. Eventually an investigation may lead to fraudulently issued policies, but the key area for fraud identification is in claims. If there is an opportunity to reduce fraud, it may well lie in the use of technology to identify claims with certain characteristics.

A: Mulvey: From the perspective of the insurance industry, fraud is no longer being viewed as a claims-only issue. Recent successes investigating large-scale ring cases have shown that fraud is an issue that crosses claims and underwriting. In these instances, coverages are secured with the intent of fraud. Misrepresentations are made at the time of purchase and the fraud is then manifest through the claim. Detection efforts during the underwriting process are being stepped up across the industry. Determining identity fraud at the time of application submission can eliminate a costly claim investigation down the line.

A: Richard Girgenti, KPMG LLP: As more insurance providers enter the e-commerce field, their vulnerability to fraud on both the claims and eligibility sides has increased. Online claims filing, policy applications and policy binding have created a new workflow and process by which insurance companies receive and review data. These new processes require greater attention to information security as confidential insurance policyholder and claimant information is transmitted electronically. On the healthcare side, there are a number of technology-based products and services intended to address HIPAA (Health Insurance Portability and Accountability Act).

Q: Are insurers devoting enough resources to fraud? Are there new tools or technologies to aid insurers in fraud prevention/detection?

A: Gummig: I'd say the answer is no. Everyone has a vested interest in deterring fraud, because it affects everyone's bottom line. Even though only about one percent of insurance sales are now through the Internet, there is a potential for $2 billion a year of Web-related insurance sales in the next few years.

I would focus on building in more upfront controls, making sure that suspicious Internet insurance activity be forwarded to the appropriate insurance fraud bureau, and continuing to build relationships with the private sector and federal authorities who have the perspective of organized, multi-jurisdictional, online fraud activity.

A: Barry: Clearly progress has been made with insurance companies cooperating with law enforcement agencies to break up large, well-organized fraud operations, but fraud persists as a major problem. The State of New York's recent implementation of encrypted bar codes for verifying proof of automobile insurance is an excellent example of using technology to fight the state's historical fraud problems.

A: Girgenti: Insurers are devoting more resources to the issue of fraud than ever before. However, with the advent of e-commerce, some traditional fraud detection and prevention techniques have become obsolete. The challenge to the industry is to develop new detection and prevention tools to keep up with the technology applications used for policy procurement, delivery and claims adjudication. For instance, bar coding is now being used as a way to store proof of insurance and history on automobile identification cards.

Q: What is happening with Special Investigative Units from a technology perspective?

A: Gummig: HR 1408 is probably the biggest thing happening in this realm. In addition, the Producer Licensing Model Act, introduced or passed by about 42 states over the past 12 to 16 months, brings about national standards with respect to licensing insurance producers. This legislation, along with HR 1408, will promote greater uniformity, and the data sharing critical to ensuring fraud is prevented or, at a minimum, detected early. The state insurance regulators have already created a prototype system which takes a digital fingerprint that conforms to FBI standards and transmits the fingerprints to the FBI, which uses it to do a criminal background check, and transmit the results back within 24 to 48 hours. Future checks could be done without the need to require new fingerprints since the digital image would already be on file.

A: Barry: The answer varies from company to company. Use of screening mechanisms or templates have to be carefully constructed to avoid "profiling" that may result in targeting a specific ethnic group or community. At the same time, scrupulous electronic examination of claim characteristics using technologies such as data mining analytics could make fraud investigation far more effective without pure manpower increases. Other technologies that ease the flow of information transfer, such as new archival and storage techniques, help make it easier for investigative units to access policy data.

A: Mulvey: SIUs at insurance carriers are developing new tools to quickly sort through large sets of data to identify commonalties that lead to detection. Ten years ago the task of insurance fraud investigation was 10 percent keyboard and 90 percent pavement. That ratio has changed to a 30 percent keyboard, 60 percent pavement split, and the trend will continue as databases mature and more efficient technology tools become available.

The ISO All Claims database is an example of great advancement because it makes 225 million loss records available at an investigator's fingertips for cross-referencing. The method of developing these tools has also changed. Third-party systems companies are now partnering with industry IT departments, and this has accelerated the process. The emerging use of data-mining and visualization functions is an example.

A: Girgenti: As a result of an increase in electronically available information, insurers are sharing more information with each other. Loss information and claims history is more readily available through various commercial databases.

-------------------------------------

THIS MONTH'S EXPERTS

GARY GUMMIG

Chief Information Officer

National Association of Insurance Commissioners (NAIC, Kansas City)

WILLIAM D. BARRY

Senior Vice President, Sales and Marketing

Docucorp International (Dallas)

THOMAS J. MULVEY

Special Investigation Director

The Prudential Insurance Company of America (Newark, NJ, more than $371 billion in assets under management)

RICHARD GIRGENTI

Principal, Forensic & Litigation Services

KPMG LLP (New York)

Peggy Bresnick Kendler has been a writer for 30 years. She has worked as an editor, publicist and school district technology coordinator. During the past decade, Bresnick Kendler has worked for UBM TechWeb on special financialservices technology-centered ... View Full Bio