The traditional ways in which insurers underwrite risks are inadequate to accurately assess and price evolving risks associated with natural catastrophes and technological innovation, according to Capgemini’s World Property and Casualty Insurance Report.

Yet insurance leaders and underwriting professionals aren’t on the same page when it comes to their confidence in the power of nontraditional tools, such as artificial intelligence and machine learning, to accomplish both, says the report, subtitled, “Become an Underwriting Trailblazer.”

The industry is plagued by rising premiums, limited provider options, insurance companies pausing new policies and dropping existing ones, and a raft of organizational constraints such as insufficient data, legacy systems and a lack of skilled talent, the study authors wrote.

Though insurance leaders remain optimistic about AI’s impact on underwriting quality and fraud reduction, underwriter confidence lags, the report found.

Insurance leaders are optimistic about AI’s impact on underwriting for increased speed and accuracy, the report found, but underwriters lack trust.

Sixty-two percent of insurance executives recognize artificial intelligence/machine learning technology (AI/ML) is elevating underwriting quality and reducing fraud. In spite of these benefits, only 43 percent of underwriters trust and regularly accept automated recommendations from predictive analytics tools. Many underwriters remain concerned about complexity (67 percent) and data integrity (59 percent).

Insurers can overcome their reluctance, according to the report, “by engaging underwriters early on to secure buy-in, retain the all-important “human in the loop” to ensure the AI/ML models are explainable and appropriately transparent and continually assess progress.”

“Today’s insurer is operating in one of the most precarious environments in recent memory. The industry must react to this volatility by rethinking the underwriting rule book,” said Adam Denninger, Global Insurance Industry Leader at Capgemini, in a statement about the report. “It requires shifting away from legacy models by modernizing core systems and deploying advanced technologies that drive better outcomes and transparency.”

Just 27 percent of insurers have advanced predictive modeling capabilities, crucial for accurately predicting potential natural catastrophe losses, Capgemini notes. A huge portion of the industry is relying on outdated models that fail to account for the speed of change in these events, both in terms of frequency and severity.

A little more than a handful (8 percent) of underwriters have successfully embraced advanced technologies, such as predictive analytics, AI and machine learning to integrate diverse data sources, including third-party data and data from connected devices, Capgemini said.

This small percentage of P/C insurers are considered underwriting “trailblazers” consistently outperforming mainstream carriers by leveraging AI-driven insights and automation to make informed decisions and accurate risk assessments with efficiency, the report found, driving greater collaboration and customer transparency by keeping underwriters at the heart of all decisions.

According to the report, trailblazers can expect to reap benefits across efficiency (higher speed and lower expenses), accuracy (loss costs and fraud detection) and customer experience (new business and policyholder retention).

The analysis finds that less than 13 percent of this group miss business goals associated with these priorities, compared to 21-36 percent of mainstream insurers.

Also reporting results of a customer survey, the report reveals that 42 percent of policyholders find the current insurance underwriting process complex and lengthy.

In search of lower premiums (60 percent) and better coverage (53 percent), 27 percent of policyholders switched carriers in the last two years.

Globally, 53 percent of policyholders express concern about the amount of personal information collected by insurers; yet, almost two-thirds say they would be willing to share more data in exchange for transparency, discounts and privacy reassurance.

Combined ratios of over 100, evolving risks due to technology innovation such as cyber threats and emergence of generative AI, and regulatory complexity have been problematic for current underwriting practices.

Industry executives cite significant organizational barriers affecting their ability to delight the customer, including insufficient access to data (54 percent), legacy systems (5 percent) and a lack of skilled talent (47 percent).

Reporting another disconnect between executive sentiment and reality, the report notes that while 83 percent of P/C insurance executives believe predictive models are critical for underwriting’s future, only 27 percent say their firm has advanced capabilities.

According to the report, while 49 percent of underwriters value drone image data yet very few insurers are equipped to support and analyze them effectively. Similarly, 1 in 2 underwriters want data from connected devices for real-time information about personal and commercial assets, although only 12 percent of insurers can capture such data effectively, the report found.

This “lack of data mastery” is detrimental to an insurer’s core business as the report found that incomplete risk evaluation plagues 77 percent of insurers.

As a result of weak data resources, 73 percent of firms face limited pricing accuracy preventing adequate claims coverage and potentially threatening solvency.

Inconsistent underwriting decisions are a prevailing issue for insurers, Capgemini reported.

The report found that only the underwriters making the best use of data, AI and automation are staying profitable.

The World Property and Casualty Insurance Report 2024 draws data from three primary sources: the 2024 Global Insurance Voice of the Customer Survey, the 2024 Global Insurance Executives’ Survey, and the 2024 Global Insurance Underwriters’ Survey. Primary research covers insights from 18 markets: Australia, Belgium, Brazil, Canada, France, Germany, Hong Kong, India, Italy, Japan, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, and the United States.