Technology is helping turn the tide against insurance fraud, but the battle against evolving threats continues
Advances and investment in technology have enabled UK insurers to be at the vanguard of the battle against fraud, yet as the war rages, a new front is opening up.
Global defence contractor BAE Systems is urging insurers to better understand the evolving threat of financial crime and by doing so provide a deterrent to would-be transgressors, as opposed to merely dealing with an incident at the point of claim.
The company’s insurance consultancy arm, BAE Applied Systems released a report which aims to help insurers address these concerns. Cate Wright, global head of insurance product at BAE Systems, explained to Insurance Times that insurers need to move from being reactive to pro-active.
“It’s about moving from the point of defence to the point of deterrent. Once a company is under attack there is a cost related to that. So, we want to work on that deterrent space. The only way to do so is by understanding the attackers,” she said.
BAE Applied Intelligence has more than 10 leading UK insurers as customers, including RSA, Hastings and Zurich, but when those who receive its data and intelligence services via the IFB are added, that rises to more than 60, which includes insurers of all sizes plus brokers and aggregators.
And while insurers have at times been accused of being slow in changing their ways, Wright claims it is one of the more innovative segments of the financial services sector, in comparison to banks, for example.
The report gives an overview of what’s going on in the UK, and also provides insights from the US.
“It’s good to look back at where we were a year ago and compare the topics that people are talking around,” Wright said.
Understanding the fraudster
The focus on fraud this year is deterrent and understanding the fraudster. This is being talked about a lot in the industry at the moment, she pointed out, with behavioural analysis a key theme.
“The industry’s in a place, particularly in the UK, where a lot of technology has been adopted. We’re comfortable addressing insurance fraud with technology now, and most UK insurers will have something in place.
“The topic now is: how do we feed that technology? Machine learning is only as good as the knowledge and learning we put in. We need to understand the environment we work in and understand the forces attacking us”, Wright stated.
“The report is about taking a step back and trying to understand who is attacking us, and making sure our technology is best-placed to stop that”, she added.
So what does a typical fraud perpetrator look like? On the non-professional criminal end of the scale, Wright pointed out that there are people who will never steal from a shop who would commit insurance fraud, for example when presented with an opportunity to inflate a claim.
“It’s hard to profile, but we’re more informed than we’ve ever been. This is due to the amount of data and information the industry now has”, she said.
“Only now we are realising that if we get really good at understanding the good customers, and that gives a better understanding of those in the ‘grey area’ [i.e, non-professional criminals].”
’Chicken and egg’ situation
Wright highlighted a ‘chicken and egg’ situation that technology creates. Technology and the increasing availability of artificial intelligence (AI) in fraud detection obviously helps make fighting the battle easier, but at the same time inadvertently creates a climate where those who would never steal from, say, a shop, are emboldened to commit fraud using the cloak of a screen and minimal contact with the victim.
“If you put a third party between the customer and the insurer (ie, aggregator) it makes the customer feel they are not defrauding the insurer directly. That barrier makes people a little more confident”, she pointed out.
“If you go back ten years when you went to an insurance broker, that was a huge deterrent. People don’t feel the guilt and the nervousness of lying to a screen. Digital broadens the horizon of people who will commit fraud”, she opined.
While AI has been happily adopted by many parts of the industry, Wright provides a caveat for its application.
“[AI] puts a lot of pressure on the insurer to make the right decision very quickly. They have a decision of whether they should pay straight away or whether they delay that process and give the customer a different journey while they investigate for fraud”, she said.
“The greatest impact AI will have is on the insurers and their customers’ experience. We know attrition is one of the greatest challenges for insurers.
“Insurers generally don’t make a profit on the first year [of a policy premium]. You need that loyalty over a three- to five-year period. If you upset that customer at point of claim because you have automated through a fraud department, you’re going to lose that customer. But the customer whose claim is paid in a matter of minutes are likely to stay with you.
“As a customer, I buy my policy online via an aggregator in three minutes. I don’t then expect my claim to take three weeks. I think automation is going to help that,” Wright added.
The notion of insurance fraud being a victimless crime is also an issue that needs to be tackled. Negative reporting in some parts of the mainstream media on issues such as dual pricing and the loyalty penalty can create a feeling of justification in the minds of some fraudsters, Wright pointed out.
“When there is a breakdown in trust against insurers, with people thinking that insurers try to avoid paying claims, they feel entitled. It’s a difficult mindset to attack”, she said.
According to Wright, the best way to address this mindset is to make people aware they are committing a crime, and helping them understand the impact on themselves and their families in the form of increased premiums.
Subscribers read more