Receive notifications whenever new stories align with your search criteria. Set up an alert to track developing stories, stay up to date on competitors, or monitor industry news. Your article has been successfully shared with the contacts you provided. There are differing opinions among stakeholders regarding new state insurance guidelines on artificial intelligence. Some individuals desire the guidelines to be flexible and accommodating, similar to a billowy wool kaftan, while others prefer them to be strict and unyielding, like a tight steel belt. These opposing viewpoints are evident in the comments surrounding a model bulletin drafted by the National Association of Insurance Commissioners' Innovation Cybersecurity and Technology Committee. The committee recently released a second draft of the bulletin on the NAIC website, with comments on the new draft due by November 6th. In an email interview, Scott Kosnoff, an insurance law specialist at Faegre Drinker, remarked that the NAIC's regulatory effort concerning AI overlaps with Colorado's new regulation that prohibits race-based discrimination resulting from the use of "external consumer data and information sources. " However, Kosnoff noted that Colorado's approach is prescriptive, whereas the NAIC bulletin outlines regulatory expectations rather than strict requirements. The underlying consensus is that life and annuity issuers should not employ AI or other emerging technologies to unfairly discriminate, although debates may initially focus on the level of flexibility and strictness within the rules rather than the goals themselves. It is important to note that the regulation of the insurance industry falls under the jurisdiction of individual states, as federal law delegates this authority. While the NAIC can establish voluntary guidelines, it typically does not have the power to impose binding rules.

The current model bulletin draft is a revised version of an earlier draft published by the Innovation Committee in July, which was included in a meeting packet distributed in August. The bulletin is part of an ongoing discussion among regulators, insurers, insurance groups, and consumer organizations regarding the utilization of new data and data analysis methodologies in the marketing, underwriting, pricing, and administration of life and annuity products. Several years ago, consumer advocate Birny Birnbaum began advocating for anti-discrimination rules pertaining to AI at NAIC events. The new NAIC draft bulletin reflects the AI principles adopted by the NAIC in 2020. The first bulletin draft generated a variety of comment letters, reflecting the questions shaping the drafting process. Sarah Wood from the Insured Retirement Institute highlighted the reality that insurers may have limited options based on the offerings of tech companies and urged the committee to approach the issue thoughtfully to prevent exclusion of compliant vendors from the industry. Scott Harrison, co-founder of the American InsurTech Council, appreciated the flexible, principles-based approach in the initial draft but proposed measures to encourage states to adopt uniform standards. Michael Conway, the insurance commissioner of Colorado, suggested that the Innovation Committee could receive support from life insurers in implementing stringent rules. The Center for Economic Justice, led by Birnbaum, argued that the first bulletin draft was too lenient and would not enhance regulators' oversight or prevent unfair discrimination resulting from AI systems. John Finston and Kaitlin Asrow, executive deputy superintendents at the New York State Department of Financial Services, supported the inclusion of stringent, data-driven fairness testing strategies, such as analyzing "adverse impact ratios" to identify any disparities in favorable outcomes between protected and control groups of consumers.