New Analysis Dispels Common Myths Behind ‘Algorithmic Bias’ Bills and Insurers’ Use of AI

NAMIC Refocuses the Discussion on Risk Mitigation

INDIANAPOLIS (Mar. 24, 2025) – The National Association of Mutual Insurance Companies today released a new analysis to dispel common myths behind widespread artificial intelligence legislation at the state level. “Big Data, Artificial Intelligence, and Risk-Based Pricing: Dispelling Five Common Myths,” discusses how targeting vaguely defined “algorithmic bias” misleads lawmakers setting rules for how insurers use AI.

This analysis comes as 18 states consider overly broad AI bills, and state insurance departments continue to adopt the National Association of Insurance Commissioners’ Model Bulletin on the Use of Artificial Intelligence Systems by Insurers. Both implement standards around the nebulous concept of algorithmic bias. The NAMIC analysis rebuts five common myths serving as the basis for this activity and outlines that:

• Insurance is a unique product that states should regulate distinctly;
• Use of AI enables widening availability of insurance to more consumers;
• Fairness for consumers means matching rates to risk, not mandating one price for all;
• Differential treatment of insurance consumers is based on risk, not protected class status;
• Outcomes-based testing methods for algorithmic bias are unreliable and incompatible with insurance regulations.

“In setting rules of the road, policymakers must recognize that insurance is distinct in function and pricing from many other consumer products,” said Lindsey Klarkowski, policy vice president for data science, AI/ML, and cybersecurity at NAMIC. “Insurance classifies based on risk and insurance law requires those risk classifications to be actuarially sound and not unfairly discriminatory.”

States currently debating flawed “high-risk” AI legislation include Arkansas, California, Connecticut, Georgia, Hawaii, Illinois, Iowa, Maryland, Massachusetts, Nebraska, Nevada, New Mexico, New York, Oklahoma, Rhode Island, Texas, Vermont, and Virginia.

Those attending this week’s NAIC Spring National Meeting will be hearing discussions on additional standards for insurer use of AI over and above those set in the model bulletin, as well as enforcement, which includes bias-testing expectations.

“The concept of algorithmic bias is incongruent with the way insurance is priced and functions, and the concept is arguably a red herring for the real underlying issue – the existence of elevated risk representation in a group,” Klarkowski said. “Rather than discussing the artificial insurance pricing, we should all be asking how best to mitigate those underlying risks that are represented in a particular risk class.”

###

NAMIC membership reflects many of the country’s largest national insurers as well as regional and local mutual insurance companies on main streets across America. NAMIC members write $383 billion in annual premiums and account for 61 percent of homeowners, 48 percent of the automobile, and 25 percent of the business insurance markets.