Insurance firms are leveraging artificial intelligence, and some envision a world in which they can predict exactly when and where a loss will occur . But as Professor Justin Sydnor points out in a recent interview with AM Best, the boundary between improving and eroding risk protection is a dangerous one to cross.
Sydnor, the American Family Insurance Distinguished Chair in Risk Management and Insurance, suggests insurance relies on a simple shared promise.
“[Insurance] works because both the insurer and the individual face a moment of uncertainty of real risk ahead of time. They don’t know for sure who’s going to be hit,” he tells AM Best, a global credit rating agency and news publisher specializing in the insurance industry.
However, better prediction modeling with AI puts that shared moment of uncertainty at risk. Sydnor likens it to using a crystal ball to predict the future.
“If those technologies get much better, and, say, with a three-or six-month lead, we have a really good prediction of which houses this year are going to be hit by a fire or which houses are going to be hit by a hailstorm, that sort of crystal ball for an insurer actually harms the insurance industry and the value proposition of insurance.”
How? If insurance models can accurately determine when and where a catastrophe is guaranteed to happen, it’s no longer a risk; it’s a certainty. Insurance becomes a moot transaction, as a family’s premium would simply equal the cost of the loss. Where an insurer created value by taking on risk, now they can only hope to trick an unaware consumer who can’t predict the future as well.
“That’s something that’s not currently being researched and thought about enough in the industry of exactly how our rapid advances in predicting losses may be eroding our ability to provide valuable risk protection,” Sydnor says.
Instead, Sydnor suggests using advancements in technology to prioritize prevention over prediction.
“The overall benefits to customers and society [occur] when these investments are not about trying to cherry pick the best risks but come from some other more fundamental improvement in the ability to mitigate, avoid, or cover risk.”
For more insights, listen to the full discussion with Sydnor:
College Professor Sydnor: How Emerging Technologies Are Redefining the Economics of Insurance