How AI plus behavioral economics are transforming insurance
By using psychological incentives, insurers now have the tools to encourage some policyholders to minimize their risks and reward them for doing so.
Innovations that will shape the future of the property and casualty insurance industry are found at the bleeding edge of technology.
Today, that edge is artificial intelligence (AI) and the data analytics used to reduce risk, and better yet, predict the future with rising precision.
While InsurTech startups have received some $1.7 billion in funding in 2016, history shows that a great majority of those companies will not succeed. While some may assume that money has been wasted, that’s not the case. Each startup tests a hypothesis in real-world conditions that anyone paying attention can benefit from. Rather than standing on the shoulders of giants, successful startups and incumbents are typically benefiting from the once nascent ideas of innovators that came before.
AI ascendant
Many of us in the insurance industry have heard the rumble in the business world about AI, as just last year, Gartner deemed AI to be at the “peak of inflated expectation.” This means it’s subject to a lot of hype but still four or five years away from mainstream adoption.
Access to stores of data via cloud computing, faster processing and cheaper storage have given a boost to decades-old AI technologies. But the real advance has been in deep learning, which creates intelligent systems by exposing them to huge amounts of information.
Gleaning the important data is the key, and AI is quickly automating most of the “grunt work.” This enables businesses to develop highly accurate predictive models to the point where such predictions will become commonplace. In the not so distant future, such predictive powers will be commonplace in the insurance industry.
At the same time, as prediction models become commoditized, there will be a renewed appreciation for judgment and persuasion. That’s where the next factor comes in.
Behavioral economics
As AI has boomed, the business world has become enamored with behavioral economics, as books written by Nobel Prize winners in Economics such as Nudge by Richard Thaler and Thinking, Fast and Slow by Daniel Kahneman have become must-reads. The premise of such books — usually rooted in rigorous research — is that the human brain is malleable and subject to predictable biases. Those who understand and anticipate those biases can help anticipate and influence human behavior.
In insurance, there’s a direct application for such research. Insurance is based on understanding risk. Traditionally, such understanding has been performed with only blunt tools, systems and measurement processes rooted in observation of past events resulting in sweeping generalizations about demographic groups.
However, by using some psychological incentives, we now have the tools to encourage some users to minimize their risks and reward them for doing so. For instance, InsurTech startup Lemonade asks its users to sign a digital pledge of honesty at the beginning of the signup process rather than at the end, and capture it using a camera instead of signing forms. MetroMile uses an in-vehicle device to judge how safely its users drive and rewards them accordingly with higher or lower premiums.
AI plus behavioral economics equals a winning combination
By studying human behavior, we can use techniques like anchoring (showing the benefit or downside upfront) and choice architecture (presenting choices in different ways to consumers) to further influence the way consumers choose to act. Establishing shared rewards among social connections can also help them achieve greater scores that result in lower premiums. An A+ consumer, for instance, can offer tips to a B+ consumer to improve their behavior, resulting in greater outcomes for the group and insurers.
AI’s role is to constantly refine such models through machine learning and study which messaging and incentives work best. Such tools are in our grasp today. The combination of intelligent behavioral incentives and AI can transform insurance from an industry based on stagnant risk assessment to one that actively tries to reduce risk. Motivated consumers benefit by taking steps to reduce the money they pay for insurance. Insurers benefit from encouraging behaviors that result in fewer payouts.
While the industry is not quite there yet, InsurTech is helping drive change and move the market closer to realizing its vision of an industry where technology delivers automation and predictive science helps insurance companies and consumers make better decisions.
Zachary Alvarez is a Strategy and Planning adviser at TransUnion. He can be reached by sending email to Zachary.Alvarez@transunion.com.
See also:
5 ways AI and data are transforming the insurance space
5 insurance and artificial intelligence predictions for 2018