Using tech to win the fight against fraud

There is nothing insurers want to do more than create the ultimate, friction-free process from end to end, for clients and for themselves. The ability to onboard customers rapidly, with all the information to hand, cuts out costly inefficiencies and promotes a fantastic customer experience—something that is essential in today’s highly competitive market.

But as many carriers have discovered, the imperative to make sure fraud is being kept at bay can significantly impede this longed-for seamlessness. It can create unnecessary delays, or allow costly fraud into the book of business, or rule out otherwise blameless customers.

On November 10, Arnaud Grapinet, chief data scientist at Shift Technology, will tell the audience at Underwriting Innovation USA how they can use data and technology to weed out fraud and improve the underwriting process to build that seamless experience.

He spoke to Intelligent Insurer ahead of the event.

What are the key complexities caused by managing fraud in the underwriting process?

Fraud can occur at the time of quote, point of sale, time of application, or point of renewal. It even happens during policy-in-force. So the first complexity is that insurers have to watch out for fraud across the policy lifecycle.

Looking for fraud at the point of quote is something you need to be doing in real time—although it isn’t necessary, for example, in a fraud investigation at point of claim. Insureds could wait a couple of weeks for an application to be processed while fraud investigations are carried out, but from a customer experience perspective, a real-time investigation creates a better outcome.

How does artificial intelligence (AI) improve the fraud detection process?

You don’t need AI for simple business rules, such as noting that someone has four cars or three claims on a policy and therefore you wouldn’t want to renew them. The issue with rules is that they’re not precise, so trying to apply them inappropriately can overwhelm the underwriting team. In particular, they can be responsible for false positives, which are a real problem and can block your portfolio growth.

That said, using AI to target something suspicious means you can look for many indicators that rules can’t detect. It’s effective in identifying fraud rings and ID theft that are difficult to find any other way. Ghost brokering and ID theft is a growing problem, so being able to deploy AI to look for it will be a boon.

What should viewers look for in emerging tools?

Ideally, the ability to detect those fraud rings and schemes. But they shouldn’t be looking for a static solution—the industry evolves very fast. Being able to integrate lots of external data sources is also key.

Insurers already leverage most of the data they have internally to price a policy, but what really helps to detect fraud is external information. For example, a bar that is masquerading as a restaurant can be uncovered as suspicious through the application of external information such as customer reviews, websites, or social media details. AI can help bring underwriters the proof of that fraud.

Fraud detection and customer experience typically don’t mix. How will this be addressed?

People would say that fraud impacts customer experience during the claims process when it prevents a payout. However, the opposite is true. Many insurers use technologies such as Shift detection to verify that a claim is valid and additional information is no longer needed.

It will be the same for underwriting. In some processes, underwriters are forced to ask for a large volume of supporting documentation, and the more you ask of a potential client, the more likely they are to abandon the purchase.

With automated fraud detection, you can have a dynamic underwriting process where fewer pieces of supporting information are needed, and the application can be concluded faster.

What do you expect attendees will take away from this session?

We are seeing a shift in the way people consider underwriting fraud. In previous years, we would see the head of underwriting blocking a fraud detection project because they felt it wasn’t their issue. Their only concern was to grow the book of business, even at the risk of including fraudulently secured policies.

There has now been a dramatic shift in priorities and there are now several synergies between claims and underwriting fraud solutions. If claims detects a large fraud ring, it will ensure anyone connected with this ring at the underwriting stage is also blocked.

It’s now about preventing fraud entering the business at any point, not only to clean up but to grow the insurer’s profitability.

On November 10 at 3:10pm Arnaud Grapinet, chief data scientist at Shift Technology, will tell the audience at Underwriting Innovation USA how they can use data and technology to weed out fraud and grow that profitability with peace of mind.

This is the biggest online event tailored specifically to underwriting—don’t miss out. Register now!