Locke Lord QuickStudy: NAIC Gets Serious About Accelerated ‎Underwriting

Locke Lord LLP
February 28, 2023

On Wednesday February 22, 2023, the National Association of Insurance Commissioners (“NAIC”) Accelerated Underwriting (A) Working Group (the “working group”) exposed for a 45-day comment period draft regulatory guidance while simultaneously exposing for a 30-day comment period a referral to the Market Conduct Examination Guidelines (D) Working Group proposing revisions to the NAIC Market Conduct Handbook.[1] The working group membership firmly asserted existing regulatory authority to regulate accelerated underwriting as activity currently occurring in the marketplace. This is a significant shift from the working group’s prior work mostly gathering information to affirmatively flexing supervisory authority over insurance carriers’ accelerated underwriting practices.

While the working group’s jurisdiction is limited to accelerated underwriting in the life and annuities marketplace, the emphasis on discrimination resulting from algorithms, data sets, machine learning, and predictive modeling applies equally to other lines of business. In fact, the legal authority the working group is relying upon, the Unfair Trade Practices Act (Model# 880),[2] is broadly applicable across lines of business, types of licensees (adjusters, carriers, producers), and types of insurance services (claims, distribution, underwriting). Other NAIC working groups with broader jurisdiction could extend these prohibitions beyond accelerated underwriting for life and annuities.


The NAIC formed the working group in 2019 to “consider the use of external data and data analytics in accelerated life underwriting.” Since then the working group has heard testimony and viewed presentations from academia, actuarial consultancies, consumer advocates, and life insurance carriers leading up to two primary deliverables: (i) an educational paper adopted by the NAIC at last year’s Spring National Meeting and (ii) regulatory guidance currently open for public comment. While the educational paper reflects the lessons gleaned over two plus years of formal inquiry in open meetings the draft regulatory guidance was developed in recent months and asserts regulatory authority under existing law. Further evidencing, the working group’s assertion of existing regulatory authority is the referral to the Market Conduct Examination Guidelines (D) Working Group to include additional guidance in the NAIC Market Regulation Handbook.

Regulatory Guidance

The regulators in attendance stressed the need for the regulatory guidance to address current market activity while asserting authority under existing state law to regulate accelerated underwriting programs in the same way as traditional underwriting programs. This assertion of broad regulatory authority is centered on the anti-discrimination provisions of the Unfair Trade Practices Act (Model# 880).[3] The draft regulatory guidance asserts that the applicable standard for review by state regulators should be whether a life insurance carrier’s accelerated underwriting programs are fair, transparent, and secure.

Fair. Whether the carrier has in place compliance and governance procedures in place to ensure data inputs are accurate and reliable, predictive models and machine learning are based on sound actuarial principles, and do not result in unfairly discriminatory outcomes.[4]

Transparent. Whether disclosure has been provided to consumers as to the use of external data in accelerated underwriting, the remedies available to consumers in the event of an adverse finding, and the ability to correct the record if there are errors in the external data.

Secure. Whether external data about life insurance applicants is properly utilized, stored, and destroyed by the carrier upon completion of the underwriting process.

Part of the fair, transparent, and secure analysis is the determination that the carrier regularly performs audits over its own compliance and governance procedures as well as auditing compliance by vendors who perform related services such as data sourcing or predictive modeling.[5]

Examples of the questions offered within the draft guidance for use by examiners include:

  • “External data sources, algorithms or predictive models are based on sound actuarial principles, including a causal or rational explanation why a rating variable is correlated to expected loss or expense, and why that correlation is consistent with the expected direction of the relationship.
  • Request a life insurer provide information about source data used as part of its accelerated underwriting programs regardless of whether the data source is provided by a third party.
  • Has the company modified its predictive models or machine learning algorithms, or the data sets used by the models and algorithms, as a result of an audit? If so, what modifications were made and why?
  • How does the company address potential unfair discrimination by ensuring the external consumer data’s correlation to risk is not outweighed by any correlation to a protected class(es).”

Peter Kochenburger of the University of Connecticut pointed out that “unfairly discriminatory” is not defined, specifically raising proxy discrimination and the extent that “unfairly discriminatory” is tied to the concept of actuarial fairness.[6] Lauren Van Buren, Chief Counsel at the Wisconsin department, responded that state laws vary but that the interpretations are informed by each state’s anti-discrimination laws as well as their insurance laws.

Market Regulation Handbook

The working group exposed for public comment a referral to the Market Conduct Examination Guidelines (D) Working Group (the “referral group”). What this means is that the working group has concluded that additional action is warranted but that action does not fall within the jurisdiction of the working group and thus another group must take on that task. The referral group will determine for itself whether the referral from the working group warrants action. Revisions to the Market Regulation Handbook would have an immediate impact as market conduct examiners use the handbook as reference when conducting market conduct exams and reviewing rate or form filings or underwriting guidelines.

What Comes Next?

Neither the working group nor the referral group are scheduled to meet during the Spring National Meeting next month. The working group will most likely reconvene in April or May to discuss the public comments received on both workstreams. The comment period for the referral regarding the Market Regulation Handbook closes on March 24 while the comment period on the regulatory guidance closes April 14. An additional public comment period would then follow meaning final approval could follow at the Summer National Meeting (in August) if not sooner.

The working group members voiced their opinion that additional model laws and regulations are warranted while urging the Innovation, Cybersecurity, and Technology (H) Committee to consider developing new models. Consumer advocate Birny Birnbaum raised the possible paradox of urging the development of new model laws and regulations while simultaneously asserting authority under existing law. The working group’s work occurs in the background of administrative action in states such as Colorado[7] and the work of the NAIC’s Privacy Protections (H) Working Group to revamp existing privacy model acts and regulations.

Locke Lord will continue to monitor for any developments. If you have questions, please contact your Locke Lord relationship partner or the authors.[8]


[1] The Market Regulation Handbook is a compilation of two NAIC reference materials, the Market Conduct Examiners Handbook and the Market Analysis Handbook. As the names suggest, these volumes provide guidance to regulatory examiners when performing market conduct examinations.

[2] Under state insurance law, Unfair Trade Practices Acts, prohibit unfair methods of competition or unfair or deceptive acts or practices, including unfair discrimination, prohibiting disparate treatment of individuals of the same class and hazard or expectation of life as to pricing, benefits, terms and conditions, or in any other manner.

[3] The Unfair Trade Practices Act (Model# 880) is not an accreditation standard and therefore states are not held to the substantially similar standard as a condition of continuing accreditation. However, Model 880 accurately reflects the consensus among the states as all state laws follow the broad contours of Model 880 and many state laws closely track the specifics of Model 880.

[4] Intention is not required when determining unfair discriminatory outcomes. In insurance law, unfair discrimination occurs when similarly situated risks are treated differently. Furthermore, disparate impact on a protected class, even without any evidence of overt discrimination or even negligence, may be deemed unfairly discriminatory, if otherwise similarly situated individuals, but for their membership in a protected class, suffer adverse outcomes.

[5] An insurance carrier can audit an algorithmic vendor without direct access to the code of the algorithm. Vendors need not risk their intellectual property to cooperate with compliance audits by insurance carriers.

[6] Actuarial fairness is the concept that the price of an insurance policy is fair if applicants bearing the same risk are charged the same price. Actuarial fairness seeks to align individual premiums with the risk an individual brings to the larger insurance pool.

[7] See,

[8] The Privacy Protections (H) Working Group plans to replace the existing Insurance Information and Privacy Protection Model Act (Model# 670) and the Privacy of Consumer Financial and Health Information Regulation (Model# 672) into one consolidated model.