Seventy percent of home insurers are either using or have an interest in using artificial intelligence in their businesses, the NAIC found.
That data was part of the second in a series of surveys state insurance regulators are conducting across the insurance world. Regulators discussed the findings during the National Association of Insurance Commissioners' summer meeting in Seattle.
An initial survey of auto insurers found that 88% of insurers currently use, plan to use or plan to explore using artificial intelligence or machine learning as part of their everyday operations. Those survey results were released earlier this year.
"These surveys were conducted to accomplish three primary goals," Vermont Insurance Commissioner Kevin Gaffney said. "To better understand the insurance industry's use and governance of big data and AL/ML. To seek information that could aid in the development of guidance or potential regulatory framework to support the insurance industry's use of big data and AL/ML. And to inform regulators as to the current and planned business practices of companies."
Gaffney is vice chairman of the Big Data and Artificial Intelligence Working Group, which heard a report on the survey Aug. 13. The survey of 194 homeowners' insurers found that they are using AI in the following ways:
54% for claims;
47% for underwriting and marketing;
42% for fraud detection;
35% for rating;
14% for loss prevention.
Reasons for not using AI
"The main reason reported for not using, not planning to use and not exploring the use of AI for homeowners' insurance were no compelling business reasons, waiting for regulatory guidance and lack of resources and expertise," Gaffney said.
Claims is a big area for use of artificial intelligence and machine learning, the surveys show.
Home and auto insurers are using AI/ML to create claim models for claim assignment decisions and settlement amounts. But the surveys show that insurers are using claim models far more for auto, Gaffney noted.
"For fraud identification both auto and homeowners' companies reported using AI mainly to refer claims for further investigation with some AI to detect organized crime rings," Gaffney said.
In rating and underwriting there was less reported use of advanced AI, he added, due to the transparency requirements of state insurance departments.
Potential next steps include exploring insurer AI model usage, Gaffney said, and evaluating the regulatory framework around the use of third-party models, and determining whether additional white papers on best practices would be useful.
InsuranceNewsNet Senior Editor John Hilton covered business and other beats in more than 20 years of daily journalism. John may be reached at [email protected]. Follow him on Twitter @INNJohnH.