IT Spotlights anti-fraud strategy: Tech tightens the net

The rise of fraudulent policy applications could become as big a headache as the already well-entrenched claims fraud, Allianz Insurance chief executive Andrew Torrance warned last month.

He called for the industry to join Allianz in fighting application fraud over the next year.

Fraud is one of the biggest problems for personal lines insurers today. And it is growing.

But insurers are now increasingly using specialist software to tackle application and claims scams, as well as relying on the traditional training and gut instinct of their claims staff.

Hi-tech tools

Insurers may have lagged behind credit card issuers and banks in using technology to cut down on fraud, but they are catching up.

Celent research in 2007 showed that few insurers used dedicated technology to identify and pursue fraud. That has begun to change. The market is shifting towards developing this software through third parties, according to Northdoor director of consultancy Rob Stavrou.

“We are seeing much more of a move towards third-party products, rather than products being developed in-house, for these security systems that sit outside the core systems,” Stavrou explains.

On the applications side, insurers can use software that sets rules on how fraudulent customer data is likely to be, alongside the more traditional red flagging used by claims staff. But the guidelines these rules engines set can be too rigid. “With any type of fraud, the fraudsters change their approach over time,” FICO’s Europe, Middle East and Africa vice-president and managing director, Mike Gordon, explains.

Because of this, many insurers also deploy more flexible analysis software that detects and flags up trends in strange customer behaviour, often the signs of fraud. Some analysis software leads to insurers catching between 60% and 70% more fraudsters, according to Gordon.

There is also a trend towards insurers using identity checking software at the applications stage. This can help to identify fraudulent behaviour, including organised criminals who launder money by buying expensive policies and then getting a refund.

“It’s a clever way of getting money into the system legitimately and then back out again,” software house Autorek director Jim Muir explains.

If a fraudulent applicant does slip through the net, insurers can still rely on anti-fraud claims technology to catch fraudsters later in the chain.

With FICO research showing that the average UK insurer receives more than 300 fraudulent claims a day, there is a lot of work for the technology to do.

Insurers use similar rules engines to the applications side when detecting fraudulent claims. However, they also rely on software that can detect links between a claimant’s information, such as a single email address being used for two different claimants. Some link analysis software lets insurers detect 20%-50% more fraud, Gordon believes.

“It is possible to identify, investigate and repudiate a fraud claim without even leaving your desk,” Admiral claims fraud operation manager Susan Evans says.

Claims databases are another useful tool. Databases, such as the Claims and Underwriting Exchange and the Credit Industry Fraud Avoidance System (CIFAS), allow insurers to share information on claims fraudsters.

CIFAS also gives insurers the tools to check if claimants have committed fraud outside the industry, such as mortgage or credit card fraud.

Covering all bases

The future of fraud detection technology will be shaped by collaboration over existing industry-wide databases, developing programs to catch more application fraud and increasing analysis of social media.

Data-sharing among insurers will become more commonplace, according to Metaskil managing director Ian Faulkner.

Celent senior analyst Catherine Stagg-Macey says insurers are also showing more interest in analysing social media, such as Facebook. For example, if two claimants say they have never met, a quick Facebook search could reveal whether they are lying.

Insurers are increasingly using technology to find fraud earlier in the policy lifecycle, Gordon says.

But one thing holding many insurers back from updating their fraud detection technology is the need to update the rest of their legacy systems first. “Most insurers want to do it, but they have too much on their plates,” Gordon adds. “Insurers with newer processing systems are more aggressive.”