Since 2012, the fraud prevention team at BGL has grown apace, with a nine-fold expansion in team numbers, and a similar explosion in the number of fraud controls in place. As a digital distributor we have more than 8 million customers across the UK and France, so fraud prevention and detection is a responsibility we take enormously seriously. It’s a trend replicated across the industry as organisations continue to build their resilience and expertise in the increasingly complex task of preventing and catching fraudsters.
Two main scams are dominating at the moment. The first, ‘cash for crash’, is nothing new. People who are known to each other, staging ‘crashes’ and then making claims. It’s a rudimentary technique, but one that has profited a number of gangs in the UK. The second is ghost broking where the element of ID theft has increased significantly in the last year. This targets the vulnerable – for example those new to the UK – by selling them insurance policies which are false. Recent scams see criminals stealing genuine financial information from places like the dark web and combining them with publicly available data to pose as genuine customers, selling on the policies to unsuspecting customers and then cancelling them before disappearing with the money. Thousands of innocent customers have been conned into buying these useless fake policies, which are worthless in the event of an accident. We know certain communities are more vulnerable than others. But of course it’s not just vulnerable people being left uncovered and uninsured: the cost of fraud pushes up premiums for all our genuine insurance customers.
There are teams up and down the country looking at fraud prevention and prosecution and, as an industry, we’re having some success. ABI figures show that in 2016 the market detected 125,000 dishonest insurance claims valued at £1.3 billion and that the level of organised fraud fell by around 30% on 2015, with 15,000 frauds valued at £174 million detected. At BGL we’ve combined complex counter-fraud measures, including the development of new algorithms at point of quote and post-sale, with a ‘back to basics’ approach including welcome postcards to new customers to help flush out ghost policies which might go undetected through a digital-only communication model.
Julie Walker-Smith, BGL Associate Director, Fraud
We’re all working with organisations such as the Insurance Fraud Bureau and police forces up and down the country, but is it time to ask whether there’s more we can team up to do?
As an industry we’re sitting on huge piles of data and if we’re really going to make it impossible for fraudsters to ply their dubious trade in the UK insurance market, it will require a more co-ordinated and concentrated effort. The benefits for all of us commercially are clear, but more important than that, do we have a social responsibility to protect customers from becoming victims of fraud? Wouldn’t it be great to think we could make the UK insurance market void of all fraudulent activity – sounds like a distant dream but there must be certain things we could implement to get us some way there. Is there a better way we can pool our intelligence more effectively to build on the commercially available databases and create a universal block list for use at point of quote? Sharing information about known fraud to prevent multiple teams from duplicating the same work?
Individually, we’re building fraud detection deeper into our propensity models, and this element of data science is critical. When combined with the diverse skills of counter-fraud teams applying their art, experience and knowledge to the data we really get results. Is there a way we could combine the expertise of our teams more effectively at an operational level? Use the skills of our people up and down the country – who are a cross between detectives and analysts – getting underneath the data and mining multiple leads in order to identify fraud?
We all know that as fast as we’re detecting new scams, the fraudsters are finding new ways to exploit consumer trends and buying habits – changing their modus operandi and moving the fraud across product lines. It’s clearly preferable for us to be able to catch fraud before it takes place – to prevent it ever occurring, but it’s a game of cat and mouse, with our teams trying to identify scams before they take hold or make it to market.
It seems to me that the best chance we have of reducing the impact of fraud across the industry is to work together, providing our teams with many more shared data sources, allowing the detectives to apply their magic touch so that we can keep up the pressure and make it harder for fraudsters. Or of course be brave and employ a reformed Frank Abagnale, à la Catch Me If You Can!