AI: The Next Big Thing in Insurance Big Data (part 2 of 2)
In an earlier post, we explored the potential impacts of artificial intelligence in big data for the insurance industry. We discussed that unlike analytics and business intelligence systems, AI isn’t just looking back, drawing conclusions from what’s already happened. It’s also looking forward – predicting, optimizing and automating what’s yet to come.
AI has other implications for insurance as well. Here are four areas where we might expect AI to cause industry disruption within the next several years.
1. The buying decision
We’ve all heard it said that insurance is sold, not bought. Brokers and agents have always been essential to bringing new customers through the door – but AI may upset that assumption. For example, thanks to AI, the macroinsurance startup Sherpa “promises to provide customers coverage … on a direct business model (no agents or commissions),” said Mark Breading, a partner at Strategy Meets Action.
2. The insurability decision
Uninsurability is the industry’s no-man’s land. When a person’s risk profile is too extreme, an insurer can’t provide them coverage without either sacrificing their own profitability or setting the premium at a price that’s not realistic for the buyer. It’s a catch-22, in which one side or the other must pay more than they can afford: the insurer or the insured.
AI could upset that balance in a couple ways. On one hand, a more granular approach to risk assessment could expose more people as uninsurable, raising the need for governmental or regulatory involvement to meet the needs of vulnerable citizens by subsidizing necessary insurances, such as health coverage.
On the other hand, however, it could do the opposite: AI could bump more people out of the uninsurable category by fine-tuning the definition of what “an insurable risk” actually is. With better and more detailed customer profiling, for example, AI could reduce the need for insurers to establish a large base of exposures to establish mutuality. And a greater ability to forecast potential loss and compute its magnitude could change how chance of loss is calculated.
In short, if AI lowers the threshold of uninsurable risk, there’s a chance it could make more people insurable, not fewer.
3. The safety decision
As AI data is shared with customers, they become more aware of their risks and habits. And when structured effectively, this information could affect the behaviors themselves – changing behavior simply by shining a light on it – which could make certain exposures more and more rare.
We’ve noted this phenomenon before in regard to driver coaching in telematics. For example, fleet owners can implement UBI driver coaching and scoring to address individual weaknesses and help their drivers improve – making the entire fleet stronger, safer and less likely to incur claims.
4. The loyalty decision
In decades past, the only real carrot that insurers could use to boost customer retention was the loyalty discount. With AI, they’ll be able to refine customer segmentation and make personalized offers, adding new options to their repertoire of reasons to stick around.
Buckle up: there’s more to come
Complexity scientist and AI technologist Francesco Corea named a slew of insurtech startups that are busy pioneering the use of AI in insurance. It’s clear that AI is already starting to make its mark on the industry. Is your policy administration system ready to manage industry disruption? If not, take a look at the adaptable, affordable solutions at Silvervine Software. Request a demo to learn more.