How to embrace AI in commercial insurance underwriting

Computer components at the San Diego Supercomputer Center at the University of California San Diego (UCSD) in San Diego, California, U.S., on Monday, March 1, 2021. MIT researchers are using the 'Comet' supercomputer to develop an artificial intelligence (AI) approach to detect electron correlation, which is vital but expensive to calculate in quantum chemistry. Photographer: Bing Guan/Bloomberg
Computer components at the San Diego Supercomputer Center at the University of California San Diego in San Diego, California, on March 1, 2021.
Bing Guan/Bloomberg

Technology has made its way into almost every aspect of human life. The number of devices connected to the Internet of Things is constantly expanding. Facial recognition unlocks your phone, social media connects you to your friends, navigation software guides you to work and autocorrect makes sure you don’t make any spelling mistakes. This digital revolution has been steady and exponential, with each addition making daily processes easier and more efficient.

Commercial insurance underwriting is currently poised to ride the wave of this technological sea change. The insurance industry has always been driven by data, and over the past few decades, our increasingly digital economy has dramatically expanded the amount of available data and resources. There were 4.66 billion active internet users around the world in January 2021, each contributing to this expanding online library. By 2025, 463 exabytes will be generated each day by people — including photos, blogs, customer reviews and social media posts. The data pool just keeps getting deeper.

More publicly available digital information means underwriters have greater access to information on potential risks. But traditional underwriting research methods are inefficient when it comes to ingesting and interpreting such huge amounts of data. Manual searches are time-consuming and open to human error and omission. Popular search engines return page after page of unrelated results to dig through. And consumer expectations create pressures to produce and return risk assessments quickly without sacrificing accuracy.

Enter Artificial Intelligence:
Artificial intelligence has been around since the 1950s, yet it continues to represent the avant-garde in high-tech forward thinking. That’s because it is constantly evolving and being applied to new disciplines. When it comes to deriving insights from massive amounts of data, AI is unmatched. Commercial insurance underwriters who leverage AI have a distinct advantage, with the ability to identify relevant information, filter out the noise, and access a complete picture of a company’s risk. With AI assistance, underwriters can instantly scan millions of online resources and access accurate underwriting insights on a business in seconds.

Business classification:
Business classification has been a longstanding challenge for commercial insurance carriers. Underwriters typically engage in arduous, manual and time-consuming processes to determine and validate a business’ classification code (e.g., NAICs code). This manual research can amount to guesswork, and often results in a single-threaded assessment of the business class based on limited data points. A single NAICS code may only reflect a portion of the services provided by a commercial insured.

AI can leverage all publicly available online resources to create a dynamic business classification and risk assessment, comprising multiple NAICS codes.

Crowd-sourced evaluation: 
By nature of doing business in the modern world, nearly every organization creates a digital trail. AI-assisted research can follow this trail to include a crowd-sourced evaluation of the business location, services, safety features and new risk insights. A video posted to social media can show if there is live music at a bar, as well as the average age of the patrons. An image added to a review can reveal if hard liquor is being served at a restaurant, or if a pool table is present. This automated process is fast and efficient, it allows underwriters to focus on risk analysis instead of time-consuming detective work.

With just a business name and address, AI data platforms can be configured to retrieve valuable data and build relevant underwriting insights within seconds.

It’s not just delivering a needle found in a haystack — it’s taking the entire haystack and spinning the straw into gold.

Efficient submission/application processing: 
Several steps in the underwriting process can be automated through AI, including data collection, data extraction, filling forms and other tedious tasks.

AI helps to streamline the submission process and increase conversion rates. Using computer vision, natural language processing, and unstructured data analysis, submission times are cut down from hours to minutes and accuracy increases dramatically. And access to an unprecedented data set allows underwriters to screen submissions that don’t match a determined risk appetite so they can focus on those that do.

Enabling the underwriter: 
AI is the next logical step in commercial insurance underwriting. It is how you tap into the data that is everywhere — and constantly expanding — and make it actionable. In just the last few years, AI has prompted a new evolution within the insurance industry. AI can help expand, deepen and interpret the range of data sources available to underwriters, and provide invaluable insights for a more thorough and more accurate depiction of risk.

More and more leading insurance carriers are using real-time data and advanced analytics to reimagine risk evaluation and enhance efficiency. In a recent PwC survey, almost two-thirds of insurers that have invested in AI report having success in using it to create a better customer experience, and almost half indicate AI helps improve decision-making.

The future of commercial underwriting belongs to those that can successfully navigate the digital realm. Because effective underwriting is based in truth, and the truth is out there.

For reprint and licensing requests for this article, click here.
Artificial intelligence Machine learning Internet of things Commercial insurance Underwriting Technology Digital Transformation
MORE FROM DIGITAL INSURANCE