The beginning of an “evolution” in how claims are handled is starting now, according to WTW’s Tom Helm
Artificial intelligence will take over the claims space within the next five years.
Tom Helm, Willis Towers Watson’s head of claims consulting, says many insurers and claims companies are only now starting their journey towards automation.
But his insight paints a picture for how new technology could be utilised within insurance claims.
“You have two levels of automation going on,” explains Helm, “process automation and smart automation.”
Process automation includes the use of robotics to replicate the functions that a handler might perform, useful for cutting out basic administration processes.
But it is in the field of smart automation, which relies on artificial technology and machine learning, where there is the most potential.
Helm said: “Smarter automation will mean more decisions are able to be made more quickly.
“It’s currently being deployed in areas like fraud detection, but it may be that for high volume low complexity claims they can create algorithms for claims automation.
“Or it may be that for more complex claims it could provide some form of decision support.
“For example it may be that this claim looks like x and here’s some case law that supports the claims handlers.”
Other uses of smart automation include in data capture and predictive modelling.
Helm added: “Historically insurers have relied on structured data – things that go into the claims administration system.
“But smarter automation offers a broader use of what can be data.
“Some of the ways we are using that is for helping insurers identify claims which may have the potential to become a large loss earlier than they are currently being identified.
“We’re picking up on the kind of language that is coming through either from the third parties’ representative or from the medical experts and we’re identifying that this has the characteristics of a claim that looks like it could be a large loss.
“This then enables the claims handling team to get that to a more technical expert, and do an evaluation, enabling them to proactively manage the claim and ensure the right investigations and rehabilitations are being deployed around the claim.
“It’s all about getting the right claims to the right people.”
Case study - Tom Helm on how Willis Towers Watson have utilised AI so far
“AI capabilities can help drive proactivity across the claims process. These are tools already becoming prevalent to help improve fraud detection; they can also predict litigation probabilities, and identify claims where the customer experience is at risk of falling below the expected standard.
“Recently our team at Willis Towers Watson has developed machine learning models for insurers, using both structured and unstructured data to identify problem claims.
“Attempting to identify these cases manually is like trying to find a needle in a haystack, but with this solution our clients can take action much earlier – providing a better customer experience and managing costs.”
Beginning the journey
He says insurers and claims companies are still at the beginning of this claims journey. Tests are currently being done by firms to understand how to utilise the technology, but the majority of firms are still handling claims in the traditional way.
“Some start-ups have taken a more aggressive approach with this and the challenge they have is they don’t have the historic data the more established insurers do,” comments Helm.
“That means it’s going to take time to get their algorithms working right as they build up more data over time.
“The more established insurers have taken a more cautionary approach as they also need to make sure they deliver on all their financial targets as well as building up their experiences around artificial intelligence at the same time.”
But Helm is confident that the challenge presented by rising customer expectations, based on what other financial industries have shown is possible through new technology, will drive progress in the claims space.
“What all claims departments are looking to do is to handle claims better, settle them faster and more efficiently,” says Helm.
“Process automation and smart automation are key enablers for claims functions to deliver on that. But the other challenge is that it is a bit of a balancing act to master.
“Yes of course you want to pay out claims a lot faster and you want to give a fantastic service to your customers, but the balance to master is to do all that while making sure you are paying the right amount on the right claims.
“That means you need claims experts who still really understand the claims and claims policy and those key aspects to ensure the right claims are being paid out at the right time.”