When its impact is considered in one word, artificial intelligence equates to efficiency. The insurance world is full of tasks that can be made more efficient, from customer service on the front end to underwriting on the back end.

So, it’s not surprising that AI is a natural fit for the insurance industry.

Eighty-two percent of insurance executives said AI is a top priority at their companies, according to a Roots survey from earlier this year. Over 90% of insurers are actively exploring, testing or using AI capabilities, the State of AI Adoption in Insurance 2025 survey found.

The integration of AI is moving so fast that regulators are struggling to keep up. To date, the main regulation produced by the National Association of Insurance Commissioners is a model bulletin on the use of AI adopted in December 2023. Twenty-four states have adopted the bulletin to date.

As former chair of the NAIC Innovation, Cybersecurity, and Technology Committee, Kathleen Birrane headed up the effort to create the bulletin. Serving as Maryland insurance commissioner at the time, Birrane has since returned to the private sector as a practicing lawyer with a focus on insurance and artificial intelligence.

Two buckets of change

Insurers are seeing two main areas of change from AI integration, Birrane told InsuranceNewsNet. First are the many ways that AI is making processes more accurate and efficient in an incremental way. Then there are the ways AI is “completely transformative,” Birrane said.

“It enables companies to do things, think about things, price things, and develop products in ways they really couldn’t do as effectively before,” she added.

While more than 90% of insurers are actively exploring, testing or using AI capabilities, only 22% of insurers have successfully deployed AI solutions in production, the Roots survey found.

Mike Upchurch is vice president of strategy for financial services and insurance at Domino Data Lab. AI is slowly making its way through the insurance work stream, he said.

“We’re seeing AI move beyond back-office automation into core insurance functions like risk pricing and claims triage,” he explained. “The true creative use isn’t just building these models; it’s the continuous, automated validation that runs beneath them. This ensures models are always responding to new market signals and, critically, don’t ‘drift’ off course and begin mispricing risk.”

Quicker and fairer

As insurers navigate the relatively early days of AI integration, they are finding massive returns. One area, in particular, is revealing the good and potential bad about AI: telematics.

Telematics refers to technology that captures real-time or frequent data about vehicle usage, driving behavior – such as speeding, braking, acceleration, cornering, time of day, mileage – location, and sometimes vehicle diagnostics.

In the commercial/fleet insurance space, 82% of insurers report using telematics — up from 65% in 2023. Over 70% of fleets reported fewer crashes and fewer claims after implementing telematics combined with driver training.

“Mobility data around how do you drive, where do you drive, and literally, how often do you brake, and all the things that are actually probably more predictive of risk are then available,” Birrane noted. “In some ways it makes auto underwriting and pricing considerably more accurate, more efficient, and more fair.”

But not without controversy. Consumers often don’t realize how much data is being gathered or who has access to it. Critics argue that telematics data may inadvertently reflect socioeconomic or geographic bias.

For example, drivers in congested or urban areas might brake or accelerate more often — metrics that could be interpreted as “riskier” behavior, even though they’re simply driving in traffic.

Legal concerns

Birrane is now practicing law at DLA Piper, where she is a partner and serves as the firm’s U.S. Insurance Regulatory Practice leader. Birrane mentions three legal things that insurance companies need to focus on with AI:

Governance and risk management controls. Good governance can protect insurance companies from costly class-action lawsuits. These AI policies continue to evolve as the rules evolve, Birrane noted, but knowing what regulators expect is crucial.

Develop a testing protocol. This is a complex and delicate area. Even states like Colorado, which passed a law in 2021 addressing the need for testing underwriting and pricing models that use external data, have struggled to identify appropriate testing protocols, Birrane said.

“You have companies that are trying to evolve responsible review and validation practices that provide meaningful results without exposing them to more legal risk than if they did nothing,” she added.

Consumer data. States are looking harder at laws on the collection, storage, and protection of consumer data, with a particular emphasis on consumer consent at the point of collection. As insurers develop AI-driven solutions, data sourcing and use, including intellectual property considerations, are important. Companies need to be aware of when and how AI development and deployment intersect with consumer data laws, Birrane said.

As integrated AI systems are developed, insurers are developing their own data, but also buying it and aggregating it from a variety of sources. Understanding the responsibility for third parties under various regulatory regimes is crucial, Birrane noted.

Jeff Wilcoxon, senior strategy and corporate development principal at VIU by HUB, said the insurance industry cannot forsake the human element when committing to AI.

“We believe that leaning into what makes us human and augmenting it with technology is what leads to the best customer experience. We’re exploring how AI can help us improve the efficiency and effectiveness of our advisors, while keeping human connection at the center of everything we do.”

© Entire contents copyright 2025 by InsuranceNewsNet.com Inc. All rights reserved. No part of this article may be reprinted without the expressed written consent from InsuranceNewsNet.com.

InsuranceNewsNet Senior Editor John Hilton has 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.