The smart path forward with artificial intelligence (AI) integration in insurance underwriting is to avoid turning AI development into a battle between humans and technology; instead, approach the relationship as a collaboration that amplifies the strengths of both.
By the end of 2024, more than 330 million people had spent a collective 2.8 billion hours – the equivalent of nearly 320,000 years – watching the first season of the hit Netflix show Squid Game.1
The Korean show features adults forced to play childhood games – with the added twist of having their lives depend on the results. The season’s fourth episode featured the remaining 80 contestants competing in a series of tug-of-war contests in which the losers were pulled off a ledge to their doom.
Is this metaphorically what is happening to humanity with the proliferation of artificial intelligence (AI) and machine learning (ML)? More pertinent to the insurance industry, are insurers setting themselves up to be dragged off a cliff by integrating AI and ML into underwriting processes?
This article examines the latter question. The answer rests in how insurance executives and technology leaders understand the possibilities and limitations of AI and ML, as well as how they foster collaboration between humans and technology.
No one can get pulled off a cliff if neither side is pulling the rope.
AI and ML have emerged as transformative forces in the insurance industry, revolutionizing traditional underwriting processes. While often used interchangeably, these technologies have distinct characteristics and applications that are reshaping risk assessment and policy pricing in different ways.
Artificial intelligence refers to computer systems performing tasks that typically require human intelligence. In the context of insurance, AI encompasses a wide range of technologies designed to mimic human cognitive functions, such as learning, problem-solving, and decision-making.
Machine learning is a subset of AI that focuses on developing algorithms and statistical models that enable computer systems to improve their performance on a specific task through experience. In underwriting, ML algorithms can analyze historical data, identify patterns, and make predictions about future risks without being explicitly programmed to do so.
Both AI and ML share the common goal of enhancing and accelerating decision-making processes in underwriting. They leverage data-driven insights to improve risk assessment accuracy, streamline operations, and personalize insurance products.
While AI and ML technology has advanced over the past 75 years – and rocketed forward in the past decade – human underwriters have for centuries been effectively making claims decisions based on increasingly advanced models and manuals. The accumulated knowledge from underwriting professionals dates back to the 17th century, when the term “underwriting” originated from the Lloyd’s of London insurance market.2
This human-knowledge dataset is massive, and it features something that AI and ML do not yet have – higher levels of reasoning and nuance based on experience.
AI and ML are already good and rapidly getting better at making underwriting decisions on standard cases – those that are objectively and easily approvable or rejectable. But vast numbers of cases fall outside of those windows, and the insurance industry has spent decades focused on building relationships with customers who want to express their unique circumstances to a person, not a machine.
Technology is not trying to pull humans into the abyss by forcing them to abandon those relationships. On the other side of the rope, insurance executives should not willingly jump off that cliff through too-rapid and irresponsible AI integration, nor should they pull technology into the void by delaying implementation. Rather than participating in a game of tug-of-war both lose by playing, technology and humans should work together in the best interest of customers.
Here are two successful applications of such human/technology collaboration.
Insurers are now integrating AI technology that turns reams of medical records, including physician notes notorious for their challenging handwriting, into actionable summaries. In the process, this technology extracts key information and highlights potential risk factors useful to underwriters as they assess an applicant’s risk.
For more challenging cases, this collaboration showcases the best of both AI and humans – technology that speeds up a laborious manual task but that still leaves room for the valuable and necessary human experience underwriters bring to the job.
The result is a significant drop in the time it takes an underwriter to review a case without sacrificing accuracy.
For example, in early 2024, RGA announced a strategic investment and exclusive global life and health reinsurance partnership with an insurance technology company that uses advanced AI to interpret and transform medical records into a comprehensive and interactive digital underwriting abstract. This technology can efficiently sift through thousands of pages in minutes and hours instead of days and weeks, leading to material time savings.
RGA is also using ML algorithms to optimize the review process. This allows RGA to streamline workflows, creating efficiencies that it then transfers to clients through supplemental underwriting programs. It also enables RGA to provide quicker cycle times on cases, leading to faster responses and increased placement.
Through this technology, RGA can handle large volumes of business, and because the system is based on machine learning, it is constantly improving by incorporating the most current information and trends.
As AI and ML technologies continue to evolve, their integration into underwriting processes presents opportunities and challenges. Navigating the complex interplay between human expertise and technological innovation is crucial.
The focus should be on partnership – collaboration in which AI/ML and humans are teammates in excellence for the customer.
How can you put AI to work for your business? Let's talk.
