Regulators adopt AI bulletin for insurers after final debate over ‘bias’

State insurance regulators made it to the finish line with a bulletin advising insurers on use of artificial intelligence and related technology by insurers.

But not without a bit of last-minute wrangling in Orlando, where the National Association of Insurance Commissioners held their fall meeting.

The Executive Committee and Plenary adopted the Model Bulletin on the Use of Algorithms, Predictive Models, and Artificial Intelligence Systems by Insurers during the event's final meeting Monday morning.

The bulletin is not a model law or a regulation. It is intended to "guide insurers to employ AI consistent with existing market conduct, corporate governance, and unfair and deceptive trade practice laws," the law firm Locke Lord explained.

"I will remind people that it is an interpretive bulletin as well," said Kathleen Birrane, Maryland insurance commissioner. "It will go to each of the individual states to consider for use and adoption in their state."

During a Friday meeting of the Innovation, Cybersecurity and Technology Committee, Birrane refereed a last-minute debate over the appearance of the word "bias" in five different places in the bulletin text. Regulators had previously removed "bias" from the definition section of the bulletin.

Still, the appearance of the word creates "an explicit encouragement to test for bias and predictive models and AI systems as well as an expectation that insurers implement documented risk management controls for bias analysis and minimization," said Lindsey Klarkowski, director of data science and artificial intelligence policy at the National Association of Mutual Insurance Companies. "The inclusion of the term bias throughout the bulletin ... creates a standard not tethered to the bulletin's governing laws."

Swap in 'unfair discrimination'

Birny Birnbaum,  executive director of the Center for Economic Justice, suggested replacing the word "bias" with "unfair discrimination" in the five instances.

But that made regulators uncomfortable.

"I almost think we might be better to just simply remove the phrase 'bias' because we don't really have agreement on what it is," said Iowa Insurance Commissioner Doug Ommen. "Recognizing how important it is to move forward with this project. Recognizing how important governance is. But also recognizing that I'm just not sure we're in a place where we can identify some of the unintentional effects."

The committee opted to vote to send the bulletin on to the Executive Committee and Plenary. Iowa abstained from the vote to do so.

Slow pace lamented

In August 2020, the NAIC adopted guiding principles on artificial intelligence after robust discussions. Regulators added language encouraging insurers to take proactive steps to avoid proxy discrimination against protected classes when using AI platforms.

The NAIC guiding principles are based on the Organization for Economic Co-operation and Development’s AI principles that have been adopted by 42 countries, including the United States.

Peter Kochenburger is a visiting professor of law at Southern University Law Center. He lamented the slow pace regulators have taken since adopting the principles three-plus years ago.

"Three-and-a-half years later, we're still no closer to sending out the specific contents, requirements or recommendations that will actually protect consumers," he said.

InsuranceNewsNet Senior Editor John Hilton covered business and other beats in more than 20 years of daily journalism. John may be reached at john.hilton@innfeedback.com. Follow him on Twitter @INNJohnH.

© Entire contents copyright 2023 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 john.hilton@innfeedback.com. Follow him on Twitter @INNJohnH.