In the past few decades, artificial intelligence has moved from science fiction to commonplace use. Unlike conventional computer programs, which simply execute a list of pre-written instructions, AI can adapt to the information provided to it. This ability allows programmers to create software that can identify patterns, extrapolate predictions and manage information more quickly and efficiently.
The ability to spot patterns and flag risks makes AI particularly valuable in commercial insurance. Both new insurtechs and established insurance companies are embracing artificial intelligence for these abilities.
In their book “Economic and Financial Knowledge-Based Processing,” researchers Louis F. Pau and Claudio Gianotti examined approximately 250 different applications of artificial intelligence to problems in various industries, including insurance. They found that the companies surveyed had turned to AI for its usefulness in solving more complex problems than traditional programs, for providing faster and more accurate information and for its potential in risk reduction.
Better Data-Driven Decision Making
For the insurance industry, AI makes it easier to manage the massive quantities of data generated by both insurers and their customers.
“Insurance has always been a business of data,” says bolt VP of Product David Lewin. “AI brings in a fresh way to interpret the data and bring meaningful insights on both a strategic and transactional level.
“Utilizing AI technologies will allow underwriters to move from manual and linear rule-based decision making to more accurate computer generated information — ultimately allowing the underwriters to automate more simpler cases and leverage the data to assist in more complex and bigger business.”
Kate Browne, underwriter’s counsel and senior claims expert at Swiss Re, underscores David’s point by noting how few industries have more data than the insurance industry does. The problem, then, is understanding what to do with all that data.
“Human beings just don’t have the ability to deal with the massive amounts of data we are getting now,” Browne says. “That’s the competitive advantage of AI, being able to use this data deluge to make better decisions.”
Thinking of AI as a Strategic Capability
The power of AI derives from its flexibility. By training artificial intelligence on particular kinds of problems, insurtechs and insurance companies can build the tools they need for a host of specific tasks, freeing up human underwriters to focus on work still beyond a computer’s ability.
“While many insurers are starting to see benefits from AI applications, the companies driving significant returns are approaching AI as a capability — not a tool,” write Sumit Taneja and Rupesh Malik in an EXL white paper.
By including six specific aspects of AI in their operational strategy — machine learning, natural language processing, behavioral data models, voice authentication, computer vision and the Internet of Things — insurance companies can maximize the benefit of AI capabilities for underwriters.
Artificial Intelligence in Action
Artificial intelligence is already changing the way insurers do business. One key but often-overlooked change is the power of AI to align the perspectives and goals of both insurer and insured.
“Historically, the parties to an insurance contract — the insurer and the insured — have always had a different set of information,” writes Denis Kessler, chairman and CEO of SCOR. As a result, both have acted strategically in an attempt to infer what the other party knew. Insurers might poll insureds about their various behaviors, for instance, while insureds might try to describe risk or loss in a way that led to greater advantage.
The distinction between what insurers knew and what insureds knew led, in turn, to a lack of trust. By using artificial intelligence both to evaluate risk and underwrite specific policies, however, insurers can align customers’ needs, understanding and interests with their own. Insurer and insured become allies in the fight against risk.
Applying AI to Risk Evaluation
Researchers have sought to apply artificial intelligence to the evaluation of risk for several years. In 2002, inventor Jill K. Jinks filed a US patent application for “a system and method for interactively evaluating a commercial insurance risk in an interactive system,” in which agents, carriers and insurance servers all played a role.
Today, insurtechs apply AI to underwriting questions in various ways. One method is by providing a more accurate view of underlying risk by accounting for factors that may not otherwise reach an underwriter’s attention. Such tools can be particularly helpful for companies seeking to improve their combined operating ratios in an era of ever-tighter margins, says Richard Hartley, CEO and cofounder at Cytora.
Yet the evaluation of risk itself is only one way artificial intelligence can help commercial insurers meet today’s challenges.
Helping Underwriters Help Customers
While AI tools and capabilities can help insurance companies better protect their own bottom lines, they can simultaneously help insurance customers receive the personalized service they demand and the coverage they actually need.
A Cognizant case study describes the use of artificial intelligence to better understand, model and predict flood risks so as to better underwrite policies and comprehend the true size and scope of the U.S. flood insurance market. By training the software on massive quantities of geographic and demographic data, Cognizant created a tool that could provide more accurate history, patterns and predictions for flooding. Such a tool can be used to write flood insurance policies more effectively, managing specific risks without overburdening homeowners with coverage that doesn’t fit their needs.
Artificial intelligence used in underwriting thus stands to benefit insurance customers as well as insurance underwriters. “Being able to consume more data automatically, we will see more customization, and customers will benefit by paying for coverage they truly need,” says Sofya Pogreb, chief operating officer at Next Insurance. This level of personalization will, in turn, increase the chances that a customer will remain loyal to their existing insurer, since the insurer demonstrates an understanding of their specific needs.
Considerations for Commercial Insurers
Artificial intelligence already plays a role in a wide range of everyday human activities. AI-based tools appear in commercial insurance workplaces, as well as in the workplaces of the companies seeking insurance coverage. The rapid spread of AI produces several new considerations for insurers.
Evaluating Risk in an AI-Driven World
When it comes to underwriting, artificial intelligence won’t just change the way insurers work. It will also change the way they evaluate risk itself, says Warren Berey, chief underwriting and operations officer at N2G Worldwide Insurance Services.
For example, businesses that use AI in their own operations, or that manufacture products with embedded AI, face risks that our understanding of traditional code-based computer operations may not account for. As technology evolves more rapidly, the window of time in which to gather information about these risks and respond via improved underwriting narrows.
Neither insurers nor cybersecurity firms take these rapid changes lightly. As AI-related risks rise, so does the insurer’s commitment to creating and implementing the technology to challenge them, says Tim Marlin, cyber product development leader at Marsh.
Other changing risks call for the implementation of AI-enabled underwriting tools as well. In addition to changes in commercial business and cyber risk driven by AI itself, property and casualty insurers currently face a world in upheaval. Pandemics, protests, a teetering economy, climate uncertainty and other challenges continue to reshape risk.
Improving the Work of Underwriting
Combined with other tools like machine learning and big data analytics, AI can help underwriters rise to the challenge. These tools “help commercial insurers bolster their capabilities to streamline the underwriting process [and] identify and price risks in an efficient fashion and thus increase profitability,” note CapGemini’s Ramesh Darbha and fellow researchers in a 2018 report.
Underwriters are frequently pressed for time. They may spend as much as 50 percent of their work time on peripheral tasks, rather than the core work of underwriting, according to a GenPact white paper.
“In a legacy environment, underwriters must prioritize new business manually, often influenced by factors other than business strategy,” notes GenPact. The result is that much new business doesn’t get written at all, even if it would represent value for the insurer. Artificial intelligence offers a way to reduce the strain on underwriters, freeing them to focus on the issues and tasks central to writing new business.
Artificial intelligence won’t replace commercial underwriters. AI does, however, provide a powerful new tool that can change and improve underwriting work. As commercial property and casualty insurance continues to face new, evolving and uncertain risks, AI offers a path forward for reliable underwriting.
Images by: puhhha/©123RF.com, Andriy Popov/©123RF.com, javiindy/©123RF.com