AI is helping insurers improve their claims processing workflows, partly by managing resources needed to
"Advisors on the distribution side are starting to use some of the available AI tools to prospect differently and to categorize their clients so that they can better tailor the products and services more quickly and really penetrate different markets," said Suzanne Grover, VP of underwriting at Coastal Wealth, a wealth management solutions provider.
Policyholders will see a difference, according to Carla Woodard, SVP of claims at Pie Insurance, an insurtech serving small businesses. "Automation can help us improve the experience by accelerating the claims process," she said. "It gets us to resolution much more quickly, and it can help us to improve transparency and communication through business rules and automated actions that can keep them informed, keep them engaged, and then drive that greater accuracy and consistency."
AI is having an impact, Woodard added, but the pace is much slower than in other industries. "The greatest impact will be felt when all of the various components of AI are put together and used to assist the claims life cycle from end to end," she said.
The slower pace of adoption may be due to caution about applying AI to adjudication of claims, according to Chris Raimondo, EY Americas insurance consulting leader. With Gen AI, for example, it's safer to use it for "synthesizing large volumes of claim information in a concise and efficient fashion for adjusters, so they're spending less time mining information in the claim," he said. "But it's still early in terms of mass adoption of Gen AI and claims."
Using large language models, like Gen AI, to make claims decisions is still in its early stages, according to Jack Baker, director of analytic solutions, Farmers Insurance.
"You have to talk about bias, fairness and lots of different things, in the context of models that frankly are complicated and not well understood even by a lot of people who are quite familiar with statistics, machine learning and those sorts of things – and you've got to bridge the gap between those," Baker said.
The first step for applying AI to claims is to create a proof of concept, according to Baker. "I'm hoping it moves quickly and that the industry is imaginative enough around this new approach to technology and models to be able to really leverage it, but that remains to be seen," he said.
