Claims leaders are increasingly turning to artificial intelligence to modernize operations, but how do they balance automation with human touch to ensure customer needs are being met?

“The predominant question I get [from customers] is, ‘Okay, if you’re using it, am I still going to be able to talk to my adjuster? Am I still going to have somebody assigned to my account? Am I still going to have somebody with experience handling these claims?” said Melissa Hill, North American head of claims at Allianz. She was speaking on a panel at Carrier Management’s annual InsurTech Summit — The CX Advantage: The InsurTech Playbook for Authentic Customer Experience.

Hill said customers want to understand what the modern claims experience looks like and how AI fits into it, so transparency is key. Customers are asking how AI will impact claims handling, how insurers will be able to respond in a crisis situation, and how the claims process will be made faster and easier.

“Everybody wants AI, but we have to be responsible with AI to know how we’re using it in order to further the claim process,” she said.

This means for many insurers, the conversation is no longer whether AI belongs in claims but how to deploy it responsibly and effectively.

“Much of the job is still administrative. It’s reading documents. It’s chasing information down. It’s documenting what happened,” said Sean Eldridge, CEO and co-founder of Crosstie. Eldridge spoke on Carrier Management’s InsurTech Summit panel alongside Hill. “And that’s not where the real value is. It’s where their time goes. And so, trying to understand that I think is really important.”

Because an adjuster’s work is so manual, an AI agent can collect information and bring it to the adjuster more efficiently. This way, the adjuster still makes the decision and uses their own skills, but AI has already sifted through the data and information about the claim.

“I call it the FBI claims handling process because they’re always trying to find stuff,” Hill said. “I’ve got to go to this file, or I’ve got to send this email, or I’ve got to go find this, and who took what picture. I mean, it’s just a lot of information that they have to consume.”

However, if more of the “FBI” elements in claims processes are being handled by AI, Hill and Eldridge agreed regulatory considerations are a challenge.

“We just have to figure out how to do it succinctly, document it, and make sure we can share it with regulators when they ask for it,” Hill said. “Because we’re highly regulated, you want to be very transparent. There’s no black box. We can’t hide anything. We have to be able to show what we did and why we did it. So, it takes a little bit longer as you navigate and work through the AI processes and where you’re trying to insert it.”

She said at Allianz, a global team is involved in the implementation of AI across various facets of the company, including claims. IT partners are used for testing to help understand where the errors might arise.

“The more you validate, the smarter [the AI] gets, the better it gets, the easier it gets,” she said. “So, it’s just real-time, ongoing governance with a tool and how it’s working.”

Eldridge added that this is an important step amid concerns about accuracy in a highly regulated environment.

“I think that there’s an opportunity for us from a change management perspective to be able to ensure that this isn’t just a black box and to ensure that transparency actually drives adoption better than any sales pitch could,” he said.

Beyond regulation, another challenge with AI adoption in claims is cost sensitivity, he said.

“Research shows that it arguably saves about 80% of the time that it would have taken that adjuster previously to review a document,” he said. “But that investment, even when it’s just a few cents a page, doesn’t show up as an immediate cost savings. It shows up as cost avoidance, so that’s a harder sell internally. And so, the question that I think more organizations are having to ask is, ‘What else could your adjusters do with that time that they’re getting back?’ Because that reframe really matters.”

He sees the use of AI in claims becoming so seamless that adjusters don’t have to think about it because it’s already integrated into their job.

“The future of claims isn’t so much copying emails and unstructured notes or spending your day reading faxes or mail that comes in,” he said. “It’s about data working seamlessly across all of these various systems, the tools, the workflows, to be able to understand what’s happening with the claim at any given point in time without all that manual work required, and making sure that you’re getting the highest and best use of the time and talents of your trained skilled adjusters.”

For this to happen, claims professionals need to embrace AI without fear, Hill said.

“Don’t be afraid of it. Embrace it because there are a lot of positive things that come out of it,” she said. “We’re helping the customer because the faster we can get to things and the more we know when a claim comes in the door, the better chance we have to serve that customer in a much more educated way. So, let’s all work together and make it work to the benefit of the customer and give that experience across the board.”