Insurance underwriting on the fast track
Artificial intelligence and cloud computing now enable straight-through processing of applications
24 Feb 2021 | Aaron Leung

With the rapid advances in artificial intelligence and cloud computing, customers and insurance agents can now enjoy an on-the-spot underwriting experience. Traditionally, an insurer requires customers to answer a long list of risk-profiling questions generated by its rules-based engine when processing applications for life insurance coverage.

The customer’s data are then transferred to the insurer’s court and verified against the risk model, designed by medical professional underwriting experts.

Throughout the process, many customers have to provide additional medical evidence for further checks and are sometimes referred to a human underwriter.

After all these hassles, in most of the cases, the customer is manually assessed, and even if finally deemed a good risk, approval still requires an adjustment to the terms and conditions of the policy.

This long and laborious process frustrates both insurance agents and customers who are getting used to straight-through processing (STP) in their banking experience, making them wonder why the insurer cannot provide the same quality of service.

With cloud computing technology becoming more available in terms of scale, cost, and efficiency, more insurance companies are striving to provide STP for their clients.

“In 2018, a life insurer would be an outlier if it was willing to move the underwriting process and customers’ data to the cloud,” says Alby van Wyk, executive vice president at Munich Re Automation Solutions. “Yet, over the past years, life insurers have become more comfortable with it, to the point that is now the norm.”

Cloud computing technology enables insurers to instantly assess the risk profile of customers. The application of artificial intelligence (AI) also refines the risk model and shortens the list of profiling questions.

By learning from past manual underwriting decisions and the outputs of rules-based engines, AI can identify the data-points in the application process which are irrelevant or redundant, thus significantly reducing the number of questions asked with little to no impact on risk taken on by the insurer.

Despite of the advantage of AI, life insurance is still lagging behind general insurance and the banking sector in adopting this technology because of the risks involved. “Assessing the risk of underwriting life insurance policies is complex,” says Wyk. “And if one makes a mistake it can be very expensive on the claim side.”

Therefore, it is important for a fintech company to minimize the additional risks stemming from the model’s error. “We estimate and quantify this risk prior to implementing AI models into insurance operations, i.e., using real-time AI decisions as part of new business underwriting or claims processes for insurers,” says Lee Sarkin, chief analytics officer, responsible for AI and advanced analytics at Munich Re for APAC, Middle East and Africa life and health business. “And, we are willing to share the risk with our clients through reinsurance, which creates alignment of interests that standalone AI consultancies cannot offer.”

Have you also read?