AI adoption in the insurance sector: views may differ – Earnix

The rise of artificial intelligence (AI) has been a major factor in conversations across seemingly almost every industry in the last few years, sometimes verging on a panacea set to revolutionise everything it touches.

Insurance has felt this brush, with a wave of technology, data and AI-driven entrants and services into the space, each promising that their product will cut red tape and inefficiency costs, garner unfound opportunities, or speed up and scale out existing business.

But there has been little evidence so far of wholesale revolution, at least in the commercial insurance sector – more like tanks amassed on a border waiting for a conflict to start.

One technology operator in this space is Earnix, which has launched an AI-driven ratings engine that works with insurers and bankers to underwrite, then price and personalise products.

The firm’s director of strategy, Aaron Wright has had a career that has seen him move from an actuarial role to one working in data science and AI.

Speaking at Earnix’s recent Excelerate 2024 conference, he told GR he sees the emerging technology landscape as one in which “the hype is still being separated from the reality”.

His comments came as he prepared to deliver a presentation at that event, called ‘Navigating Insurance Industry Changes: From Flip Phones to Smartphones’.

“I recently had a conversation, and they said that they didn’t understand what AI is,” Wright (pictured) says.

“Most of the time, that’s the problem, which is that people’s knowledge of AI starts with what they’ve seen on television. Quite often, they think it’s about robots taking over the world. They’ve seen The Terminator, and they know how it ends!”

Wright’s definition of AI is more to the point.

“For me, it’s about using computers to do the things that people have done in the past,” he says. “It’s taking machine learning and making it operational or democratising it. You’re just using computers to do things better and with more efficiency than they’ve been done in the past.”

One of the key areas for the use of AI in the insurance world has been around managing underwriting rules and improving efficiencies. Mass market personal lines are seeing the benefits first, almost inevitably.

Inflationary costs of manufactured parts, such as in motor business, is an issue that highlights one major way AI can help, he suggests.

“One company I worked with tried to examine the cost of a Ford F-150 [pickup truck] when they switched the bodies so that they were made from aluminium,” he says.

“The problem was that they no longer knew what to charge because they could no longer look at their history to determine the price,” Wright continues.

“If you can’t do that to keep profitability where you need it to be, then you can have rules management in place to hold off while you collect data and can the synchronise it with the pricing algorithms you have in place,” he says.

As to what place AI will occupy in the future, Wright observes that there has been a marked increase year-on-year in the reporting of its usage. That proportion, however, comes from a small base.

“We asked people how much of an impact that AI was having on operational decision, and what those decisions were,” he says.

“That would be things like pricing and underwriting risk, claims, reserving, fraud detection, etc… The story throughout was pretty much the same,” he continues.

“For the previous year, they said it was that 5% of decisions were influenced by AI. This year, it was on track to be 10%, with the prediction that it would triple for next year. Their overall target is that a third of decisions will be done this way. So, they’re at least bullish about finding ways that AI can bring value,” he adds.