From the April 2009 issue of Claims Magazine •Subscribe!

The Whole Truth

Fraud losses have been an obstacle for the insurance industry since its establishment. To help detect and deter it, there is now a frequency-based technology that determines veracity by analyzing brain activity conveyed through the voice, not through physiological response (i.e. computer voice stress analyzer, polygraph machines). In England and the U.S. alike, this technology -- referred to as layered voice analysis -- has been a very useful tool in pinpointing, investigating, and resolving fraudulent claims in the insurance industry. As an investigative focus tool, it has also been a success in U.S. law enforcement, corrections, and federal government agencies.

Through its use as an investigative focus tool, layered voice analysis assists investigations by guiding users toward the development of effective lines of questioning. It gives the user the ability to determine, in real time, the subject's state of mind and emotional status. Throughout the interview, variables reflected in the subject's voice -- such as stress levels, cognitive processes, and emotional reactions -- are identified. Investigators can often use these variables to decide on topics to further explore, many times forming lines of questioning that result in confessions. Layered voice analysis has had success in many types of investigations, including special investigations unit interrogations, criminal investigations, gang investigations, military intelligence, and internal investigations.

Appropriate Application

While its use in insurance investigations is relatively new in the U.S., layered voice analysis' use in law enforcement is fairly well established. Reporting a fraudulent claim is a criminal activity, and the reactions to interrogation are closely related to the reactions of more mainstream criminals. Both types of criminals face a high level of cognitive dissonance when facing this fraud-fighting tool. In fraudulent claim investigations, the claimant will even retract a claim when he feels he is going to be discovered. In England, the technology has also served as an effective deterrent, as criminals know they will encounter it.

The use of this technology in the insurance industry is becoming more widespread and has had success. A London organization, the Harrow Council, has been using it in their claim and fraud units and has accomplished a lot of success. Because of it, the British government plans to implement layered voice analysis in organizations across the country to fight fraud. One large auto insurance company in the U.S. has also used the technology with marked success, implementing it in SIU investigations of false claims.

The success of this unique tool as a deterrent is based on its unique nature. Technologies such as the computer voice stress analyzer (CVSA) and polygraph machines are often compared to it, despite the fact that they are fundamentally different. Both CVSA and polygraphs measure stress in the voice, and determine from that stress whether the subject is lying or not. Both of these technologies work by using closed-ended questions that require "yes" or "no" answers. Many factors not related to deception can induce stress, making the determination of truthfulness based purely on stress very unreliable.

Layered voice analysis is a tool used in a different manner than the other technologies -- and it achieves a different end result. While CVSA and polygraphs measure physiological stress, layered voice analysis measures many different psychological attributes. It is based on brain activity that is captured through the human voice box, which is used as the medium, displaying the attributes with the frequencies that are put out. This brain activity is separated into 129 different parameters determined using innovative mathematical algorithms.

These parameters point the trained user in the direction of the deception. This tool is used with open-ended questions and can be used over the phone or in a live interview. The interview can then be saved as a recording and separated into relevant segments. These segments are then put into offline mode, where a deeper analysis is done using the information deemed as vital. It also allows the user to re-evaluate the interview to form more significant lines of questioning for another interview. This process lets investigators pinpoint areas of concern, allowing them to discover new areas to investigate or reach a decision that the subject has no guilty knowledge.

Comparison Shopping

In recent years, there have been numerous studies on the accuracy of layered voice analysis, CVSA, and polygraph machines. One study is currently near completion. Dr. Michael Adler conducted this double-blind study on the use of polygraph and layered voice analysis on convicted criminal subjects. Each tool was used according to the manufacturer's instructions. The abstract of the study reveals the ability of the latter to pick up on the subject's deceptions with no inconclusive results. In contrast, the polygraph had several inconclusive results. An important factor in this study was the ability of the investigator to pick up on many issues revealed in the readings that were not revealed by any other technology.

There are few tools -- outside of intuition and investigative prowess -- available to assess risk and minimize the amount of money paid out because of fraudulent claims. This is the only technology that can be used over the phone or in person in real time. It can be used in any language by both claim specialists and investigators in the special investigations units. It reconfirms what the specialists and investigators may already suspect, adding an additional red flag to a suspicious claim.

In today's economy, it is becoming increasingly important to eliminate wasteful spending. While insurance companies cannot afford to lose money to fraud, they also cannot afford to withhold money from those with valid claims. With the use of smart, innovative technologies, it is possible to assess risk, pay legitimate claims in a timely fashion, and dramatically reduce the money spent on fraudulent claims.

Page 1 of 2
Comments

Resource Center

View All »

Complimentary Case Study: Helping achieve your financial goals By:...

Find out how a Special Investigation Union used TLOxp to save the company money and...

Do Your Clients Hold The Right CDL License?

Learn about the various classes of CDL Licenses and the industries that are impacted by...

Integrated Content & Communications: A Key Business Issue For Insurers

Insurers are renewing their focus on top line growth, and many are learning that growth...

High Risk Insurance Coverage in the E&S Market

Experts discuss market conditions, trends and projected growth in a rapidly changing niche.

Top E-Signature Security Requirements

This white paper covers the most important security features to look for when evaluating e-signatures...

EPLI Programs Crafted Just For Your Clients

Bring us your restaurant clients, associations and other groups and we’ll help you win more...

Is It Time To Step Up And Own An Agency?

Download this eBook for insight on how to determine if owning an agency is right...

Claims - The Good The Bad And The Ugly

Fraudulent claims cost the industry and the public thousands of dollars in losses. This article...

Leveraging BI for Improved Claims Performance and Results

If claims organizations do not avail themselves of the latest business intelligence (BI) tools, they...

Top 10 Legal Requirements for E-Signatures in Insurance

Want to make sure you’ve covered all your bases when adopting e-signatures? Learn how to...

Claims Connection eNewsletter

Breaking news on disasters, fraud, legal trends, technology, and CE initiatives for the P&C claim professional – FREE. Sign Up Now!

Claims-Handling Guidelines

Claims Magazine is providing the following free guidelines and regulations in order to help adjusting professionals stay abreast of each state’s unique property and casualty claim-handling requirements.

View our State Guidelines »

Advertisement. Closing in 15 seconds.