28 September 2022

2022: Q4 Trends in AI for Insurance

Insurance has stood steadfast for centuries as time and technology have changed around it. Today, the insurance industry is changing rapidly in response to the ongoing digital revolution. AI is one form of digital disruption taking the insurance industry by storm.

“Artificial intelligence (AI) has the ability to enhance the insurance industry’s value chain by altering relationships, reinventing business platforms, and expanding hidden data,” write Kevin H. Kelley and fellow authors in a 2018 article in the Risk Management and Insurance Review.

Many promises have circulated about AI’s revolutionary abilities for insurance. The technology’s actual progress has been slower — but artificial intelligence is in fact changing how insurers approach distribution and customer relationships.

Artificial Intelligence and Insurance Distribution

Artificial intelligence didn’t suddenly appear on the insurance scene in 2022. Rather, its growth as an insurance tool has paralleled the increasing reliance on data and digital technologies in insurance, in related industries and in customers’ daily lives.

“AI is growing due to the ever-increasing ‘datafication’ of business interactions, private life, and public life,” researchers Naman Kumar, Jayant Dev Srivastava and Harshit Bisht wrote in 2017. Five years later, the data created by human business and personal endeavors continues to explode — and so does the need for, and use of, artificial intelligence.

Artificial intelligence has developed over the past few decades into several more focused tools and applications. AI, along with machine learning (ML) and deep learning (DL), offer three particularly powerful ways to improve insurance distribution, write Ramnath Balasubramanian, Ari Libarikian and Doug McElhaney at McKinsey:

Artificial intelligence describes all the ways in which machines mimic human cognitive functions, like learning, spotting patterns and making predictions.
Machine learning focuses on AI’s learning-like abilities to train algorithms to sort and respond to data inputs in various ways. ML can also be tasked with making predictions based on existing data.
Deep learning further distills ML’s abilities to create abstractions from data models. DL often uses neural networks to copy brain functions by connecting various niche AI algorithms together.

In day-to-day insurance distribution, these tools may appear in a wide variety of forms. One example is in communication tools used to connect with customers. AI-enabled chatbots offer one way to support value creation by cultivating customer relationships, explain Mikko Riikkinen and fellow researchers in an article in the International Journal of Bank Marketing.

Artificial intelligence can adapt to customers’ word choices and the content of their responses, allowing the chatbot to gather information quickly and naturally. The chatbot can then pass this information to an agent or customer service representative to continue the conversation.

Machine learning and deep learning also offer new ways to analyze risk and loss data, model possibilities and likelihoods, and approach underwriting. As underwriters incorporate this data for more nuanced decision-making, new areas of loss prevention and customer protection open up.

Where Is AI Headed Next?

Research in artificial intelligence suggests not only that AI provides value in present situations or in examining past datasets but also that AI can make predictions about future events based on past information.

Insurance has long been in the business of addressing potential risks based on information about previous losses. Yet for centuries, the primary focus has remained on loss compensation.

As AI becomes more common in insurance, “the insurance business model will shift from loss compensation to loss prediction and prevention,” write Martin Eling, Davide Nuessle and Julian Staubli in a 2021 article in The Geneva Papers on Risk and Insurance. As the industry’s focus shifts, insurers will discover new ways to reduce costs while maintaining or increasing value for customers.

A focus on loss prevention includes a focus on customer behavior. AI also offers other ways to analyze and understand customer decision-making.

A 2020 study published in Risks, for instance, describes an AI model “that can be employed in order to explain why a customer buys or abandons a non-life insurance coverage,” write researchers Alex Gramegna and Paolo Giudici. The model can gather information about customer behavior in real time, allowing agents and customer service representatives to better understand and address customers’ needs in the moment.

Many insurance companies already report that AI is paying off for them. A PwC survey of insurance executives finds that:

65 percent say AI is already helping them create better customer experiences.
49 percent credit AI with improving internal decision-making.
47 percent report improved efficiency, productivity or cost savings from AI use.

Even among insurance executives who aren’t currently seeing benefits, enthusiasm for AI remains high, write PwC’s Matt Adams, Marie Carr and Anand Rao. For instance, although 47 percent of respondents say they hadn’t yet seen revenues increase from AI use, they expect to see these results within two years. Insurance companies remain committed to investing in artificial intelligence, even when the payoff is not immediate.

Incorporating Today’s AI for Tomorrow’s Applications

Despite media speculation about a future driven entirely by machines, today’s AI models and algorithms can’t make their own decisions, writes Adam Uzialko at Business News Daily. Artificial intelligence can’t replace human agents, underwriters or other decision-makers.

What the technology can do is to provide summaries of datasets and insights gleaned from that data to human decision-makers. An agent or customer service representative can review that information and respond to specific customer needs. AI can also reduce the time that human professionals spend on menial or repetitive tasks like copying a customer’s name and address onto various forms.

“The result is insurers who are better equipped to sell customers the plans most suited for them,” writes Uzialko.

PwC’s Scott Likens, Michael Shehab and Anand Rao write that insurers will need to focus on several areas to ensure effective, responsible implementation of artificial intelligence in the coming years:

Assess risks and make a plan to monitor risk. Insurers will need to consider how AI interacts with their financial, operational and reputational opportunities, risks and practices.
Never “set it and forget it.” AI learns and adapts based on the information it can access. As a result, humans must monitor AI’s adaptations and output to reduce the impact of unacceptable bias and to guide AI development.
Make ethics an essential part of AI. AI can learn and adapt, but it doesn’t have the store of experiences and human interactions that humans take for granted when we learn new things. Treating ethics training as an essential part of AI use is a must.

Teaching insurance professionals to work alongside AI is also essential for insurers that wish to keep growing beyond 2022. Team members will need to adapt their own workflows. They will also need to understand how AI works, what its limitations are and where it needs to be monitored.

Once the stuff of science fiction, artificial intelligence is now a daily reality for insurance carriers, agents, brokers and customers. As the technology grows and matures, it stands to change how insurers do business — and how they provide value to customers while protecting their own bottom lines.

Images by: rh2010/©123RF.com, marchmeena/©123RF.com, sebra/©123RF.com

News