IAG treats generative AI as an “inclusive innovation” – Financial Services – Marketing – Software


Insurer IAG is using generative AI for “inclusive innovation”, so far reducing claims review times by 60 percent and enhancing the development of marketing materials.



IAG chief operating officer, Neil Morgan

At an investor day, IAG chief operating officer Neil Morgan outlined the insurance group’s “inclusive” approach to AI, seeing it as beneficial at a whole-of-enterprise level.

The focus, therefore, is on getting AI tools “into the hands of everyone, right across our business.”

“The timing of generative AI becoming ubiquitous and us completing our retail platform transformation is a huge opportunity for us. Our ability to adopt these tools and innovate on top of the core is very different to where we were three years ago,” he said.

Morgan said that generative AI uses fit under three categories – deploy, shape and compose – with investments organised around this.

“With deploy, the focus is to create every bit of value possible through the use of the capability that’s embedded in enterprise applications,” he said.

“The most familiar [deploy tooling] is Microsoft Copilot, but we’re working with many of our partners to understand and influence the product roadmap for how those capabilities will be deployed in their platforms that we can then get into the hands of our full workforce.”

Other partners also include New Relic, GitHub, NICE CXone and Appian.

“The second category is where we’ve established a bespoke GenAI bot-builder called ‘Geni’. The distinction here is that it allows us to support IAG-specific use cases, so focused on IAG information, built rapidly at low cost, addressing specific business problems or opportunities.”

Morgan said IAG is “driving a highly inclusive approach” to AI adoption by “enabling employees right across the business to benefit from this category of capability.”

“To do that, we’ve trained 150 ‘activators’ to support our people in building these solutions,” Morgan said.

“They’re across all divisions and those ‘activators’ can advise and select the right backend models, consider responsible use principles, and use low-code and no-code solutions to build these solutions out.”

In its ‘compose’ category Morgan said IAG is investing in “deep AI capabilities” before scaling them across the company.

“These are built by our specialist team creating custom solutions, and we’ve seen some good wins in all of these categories.

“Just to give you a feel for scale, we have over 100 use cases now in experimentation or trial and production.

“The teams have used embedded tools to reduce the review effort on claims correspondence – not by a little bit, by 60 percent.”

The company has also built an IAG-specific solution to “change the tone of voice across a whole swathe of marketing material – and that’s saved us hundreds of thousands of dollars.”

Morgan added, “We’ve also had a big win with delivering a claims concierge, which has about 500 active users.”

Transformation efforts

Outside of AI adoption, Morgan highlighted IAG’s transformation work around claims, which involved creating a new consolidated foundation “on which we could quickly and easily launch new features for the entire organisation, including claims tracking and digital lodgement improvements.”

“We’ve prioritised operational efficiency, better informing customers of their claims progress and getting customers back on the road and back into their homes faster,” he said.

Morgan added that transformation works also covered retail distribution, pricing and policy management across its various businesses.

This work has also allowed IAG to decommission some older systems.

“Today, we’ve decommissioned almost 300 assets, 289 to be precise,” he said. “There’s more on the roadmap as we complete our transformation streams.”

Morgan also highlighted ongoing changes associated with the re-platforming of IAG’s retail businesses onto a single technology stack, noting that new features could now be built once and deployed to all brands.



Source link