Gen AI reveals underwriting, customer service applications for insurers, Nationwide and others say

Many insurers are already pursuing their own AI “centers of excellence,” budgeting and planning to begin building in 2024, states Cindy MacFarlane, global insurance, AI and data leader at Deloitte.

Cindy MacFarlane, global insurance, AI and data leader at Deloitte.

Gen AI may have spurred these actions. “We’re seeing this really become a reality compared to a couple years ago where it was – AI, nobody really wanted to talk about it,” MacFarlane said. 

Insurers are paying more attention to AI and looking at how to harness Gen AI for both property and casualty risk, underwriting and customer service functions, as well as internal functions such as workflow management. 

Gen AI can change the internal workings of an insurer by taking over simple tasks so associates are free to focus on serving policyholders, as Jim Fowler, chief technology officer of Nationwide states. “We see Gen AI as using artificial intelligence to create virtual assistants, virtual workers, that will automate the simple tasks that our associates do, to free them up to do what we believe they do best, which is provide judgment, to be empathetic to members and customers, and to bring critical reasoning and thinking skills to every situation,” he said.

Jim Fowler, chief technology officer, Nationwide.

Also, Gen AI can increase the value and usefulness of data, according to Fowler. “We’ve got this foundational set of transactional systems that have created a single source of truth when it comes to data, to allow us to enter into what we think is a new phase for us, which is what we’re calling our augmented intelligence phase where we’re going to empower a bionic enterprise that brings humans plus machines together, to be able to run the company and serve our intermediaries, our members and our customers in a very different way,” he said. 

The complexity of insurance policies, in both life and property and casualty, could be better communicated by using Gen AI, according to Kimberly Harris-Ferrante, distinguished vice president analyst, Gartner.

“Instead of having a person searching through your policy system, or going through and reading an actual policy, you could generate a summary of that using natural language that you then either give to a human, whether it’s a claims adjuster or a call center rep, to say here’s what the coverage is,” she said.

Kimberly Harris-Ferrante, distinguished vice president analyst, Gartner

The technology can fulfill several functional points for insurers, as Harris-Ferrante explained:

P&C claims adjusters can use Gen AI to get a summary of what is and is not covered in a policy as they respond to a claim.In commercial underwriting, Gen AI could summarize the risks of a specific property. Gen AI can combine service chatbots with analysis of the customers they serve to determine customers’ eligibility for policies.Gen AI can summarize legal and compliance documents to more clearly show what an insurer has done to comply with regulation.Gen AI can be used to develop new insurance products and generate marketing messages.Pairing Gen AI with cloud data can generate sentiment analysis of customers that service representatives can use for a personalized approach to the customer. Similarly, personalizing marketing messages based on sentiment analysis increases their rate of return. 

Gen AI also has the potential, according to Harris-Ferrante, to personalize P&C and life insurance underwriting. 
“It’s in the moment of data entering a system. It’s in the moment of the conversation with the customer. It’s in the moment an alert comes in from [an] IoT device and you can generate a message to the customer that says, ‘Hey, this data just came in, are you okay?'” she said. “It’s generating some type of thing that generally is in the moment, highly personalized and highly contextualized to the situation because it has that in the moment queue and in the moment data.”

While insurers may not yet be applying Gen AI to these external, customer-facing functions, Deloitte’s MacFarlane believes that while “some are all in on use cases that are internally focused, we know that this is going to go more external, when the regulatory bodies have an idea of how they want to regulate this and we can track the models better and understand hallucinations and false positives,” she said. “It’s only the beginning. We’re seeing good traction there.”