IAG lifts lid on CASI, a new AI claims assistant

It's designed for the intermediated space

IAG lifts lid on CASI, a new AI claims assistant

Claims

By Daniel Wood

Insurance Australia Group (IAG) is scaling up its use of Generative AI (Gen AI) in the intermediated space. One big change since 2024 at CGU, the giant insurer’s intermediated brand, is the use of AI technology in the claims area. Claims Assistant Supporting Intermediated (CASI) was developed to help the firm improve how it deals with claims from brokers’ customers.

“That's rolled out to 500 or so front end claims people in intermediated and it's proven to be so powerful that now it's actually going out across other divisions as well and other claims teams,” said Neil Morgan (pictured above), IAG’s chief operating officer, in a recent interview with Insurance Business.

“Australia is a good example where our intermediated team has activators within their world and they're looking at opportunities to use these tools to optimise things,” said Morgan.

He said one of CASI’s roles is to help claims handlers find relevant terms in complex PDS documents. When it does that, he said, it also provides references so a handler can trace where the information came from.

“That's taken that process from 20 to 30 minutes [duration] to seconds,” he said. “It’s really speeding up our ability to respond to specific questions on coverage terms or specific details in a PDS - and we'll keep building it up from there.”

Morgan said CGU started scaling up the use of CASI during the second half of 2024, which suggests brokers should now be noticing claims handling improvements.

Meet CASI – the AI Claims Assistant

In November, a webinar hosted by the National Insurance Brokers Association (NIBA) and CGU provided more CASI details. In “Meet Casi – the AI Claims Assistant,” Tim Rafton explained how this Gen AI tech is one important part of CGU’s significant updating of legacy systems.

He said the focus in the claims area, ahead of making tech changes like CASI, was understanding what needed to change in work practices so the tech could really help improve claims handling.

Two big problems for CASI

A major reason for developing CASI, he said, was to deal with two issues reported by staff that became apparent early last year following broader operational changes in the claims area.

Rafton said, because of these operational changes, his claims handlers were needing to deal with an increased number of different PDS documents – up to 1,000 different types across the business.

This coincided, he said, with a changed work environment that meant many handlers were working from home.

“In a work from home environment, our traditional learning models are no longer as usable,” said Rafton.

In years past, junior staff working in an office environment could easily question a more experienced colleague sitting nearby when they had a claims question or issue. Remote working has significantly reduced that on-the-job learning.

“That’s where we came up with, ‘how do we solve that problem?’ And we started to turn to generative AI,” said Rafton.

Morgan said CASI is like a simple to operate app sitting on a claims handler’s computer. He said the handler can ask it questions verbally in natural language and from its training on IAG’s content it can find answers from the data store supporting all of the firm’s PDS documents.

“You can ask it very specific questions about very specific examples and points of reference and then it comes back with its findings and allows you to verify that finding or get more information through the tool,” he said.

Morgan said the experience of developing CASI has shown him the ample potential for AI technology use across different areas of IAG’s business.

Claims handling in the government spotlight

In October, the government’s inquiry into insurers’ response to the 2022 floods handed down 86 recommendations. Many of the recommendations focused on claims handling.

For example, recommendation 51 called on the Australian Securities and Investments Commission (ASIC) to develop key outcomes measures for the consumer experience during a claims process. Among the measures: claims acceptance rates, average response time to a customer once a claim is started and “Overall customer satisfaction” and “Consumer comprehension of communications.”

Are you an insurance broker who works with CGU? Have you noticed improvements in the claims handling process? Please tell us below.

Related Stories

Keep up with the latest news and events

Join our mailing list, it’s free!