Insurance Curator DENVER — Life insurance companies operating in Colorado are facing a pivotal deadline this month as the state’s Division of Insurance (DOI) intensifies its oversight of artificial intelligence and algorithmic underwriting. The implementation of Regulation 10-1-1, which stems from landmark legislation passed in 2021, marks the first time a U.S. state has required insurers to actively prove that their automated systems do not produce discriminatory outcomes based on race or ethnicity.
The regulations, which reached a critical reporting milestone in early 2026, require life insurers to submit detailed governance and risk management frameworks. These reports must outline how the companies monitor their "Big Data" sources and predictive models for "unfair discrimination." The move has set a national precedent, prompting a scramble for compliance among carriers who have increasingly relied on algorithmic tools to speed up the application process and bypass traditional medical exams.
“Our goal is not to stifle innovation, but to ensure that as the industry moves toward high-speed, automated decisions, the historic biases that have plagued financial services are not baked into the code,” said Colorado Insurance Commissioner Michael Conway in a recent briefing regarding the implementation. “Transparency is the only way to maintain consumer trust in an AI-driven market.”
The Scope of Compliance
The compliance requirements are tiered, focusing on two primary areas: governance and testing. Under the current rules, any life insurer using external consumer data and information sources (ECDIS)—which can include everything from credit scores and social media activity to retail purchase history and geographic data—must establish a rigorous internal oversight structure.
Insurers were required to submit their first comprehensive governance reports by the end of 2025, detailing the specific algorithms in use and the personnel responsible for their oversight. As of February 2026, the DOI has begun the "testing phase," where companies must provide quantitative evidence that their models do not result in a disproportionately negative impact on protected classes.
Specifically, the regulation mandates:
- Inventory Management: A complete listing of all predictive models and ECDIS used in underwriting.
- Bias Testing: Mathematical audits to detect "proxy discrimination," where seemingly neutral data points (like zip codes or educational attainment) correlate too closely with race.
- Third-Party Accountability: Requirements for insurers to vet third-party vendors who provide the data or the AI platforms themselves.
Industry Pushback and Technical Hurdles
The transition has not been without friction. The American Council of Life Insurers (ACLI) and other trade groups have expressed concerns regarding the technical feasibility of the testing requirements and the protection of intellectual property.
“The challenge for many carriers is not a lack of will, but the sheer complexity of the data,” said Sarah James, a senior compliance consultant at a national insurance law firm. “Colorado is asking for a level of transparency that often butts up against trade secret protections. Insurers are worried that disclosing the inner workings of their proprietary models could give competitors an unfair advantage.”
Furthermore, the "testing" component of the rule requires insurers to collect or estimate the race and ethnicity of applicants—data that life insurers have historically been prohibited from collecting to avoid discrimination. To comply with the new rules, many companies are now using Bayesian Improved Surname Geocoding (BISG) or other statistical methods to estimate the demographics of their policyholder base for auditing purposes.
“It creates a bit of a paradox,” James added. “To prove you aren't discriminating based on race, you suddenly have to find a way to track the race of everyone you interact with.”
A National Blueprint
Colorado’s Senate Bill 21-169, which authorized these rules, is being closely watched by the National Association of Insurance Commissioners (NAIC) and regulators in states like New York, California, and Connecticut. While many states have issued general "bulletins" regarding AI ethics, Colorado is the first to codify specific reporting and testing requirements into law.
Data from the DOI suggests that approximately 120 life insurance carriers are currently subject to the reporting rules in the state. Early reviews of the 2025 governance filings indicate a wide variance in readiness. While some large-scale national carriers have dedicated "AI Ethics" departments, smaller regional players have struggled to meet the technical documentation standards.
“We are seeing a bifurcated market,” said Robert Miller, a researcher at the Center for Insurance Policy and Research. “The larger companies are viewing this as a cost of doing business and are building internal auditing tools. The smaller companies are looking toward 'compliance-as-a-service' vendors to help them navigate the Colorado requirements.”
Consumer Protection and Market Impact
For consumers, the rules promise a "black box" no longer remains entirely opaque. Proponents of the regulation argue that algorithmic underwriting, while fast, has the potential to exclude minority communities from affordable coverage if left unchecked.
“If an algorithm decides you are a higher risk because you shop at certain stores or live in a certain neighborhood, and those factors are just stand-ins for race, that is a violation of the law,” said Martha Campbell, a consumer advocate with the Colorado Consumer Health Initiative. “These rules put the burden of proof back on the multi-billion dollar companies, where it belongs.”
However, some industry analysts warn that the cost of compliance could lead to a slight increase in premiums or a reduction in the use of "accelerated underwriting"—the very technology that allows people to get life insurance in minutes rather than weeks.
“There is a balance to be struck,” said Commissioner Conway. “We want the 15-minute life insurance policy to exist. We just want to make sure it’s a 15-minute policy that is fair for everyone, regardless of their background.”
Looking Ahead
As the February reporting window closes, the Colorado DOI is expected to issue its first public report on the state of AI in the life insurance industry by mid-2026. This report will likely highlight common pitfalls found in the initial filings and may lead to further refinements of the rule.
The next frontier for Colorado regulators is the expansion of these rules into other lines of insurance. The DOI has already begun preliminary workshops to apply similar algorithmic transparency requirements to private passenger auto insurance and homeowners' insurance, signaling that the era of unregulated AI in the insurance sector is coming to an end.
For now, the life insurance industry remains the "test case" for a new era of digital accountability. As the first state to move from theory to enforcement, Colorado’s success or failure in implementing Regulation 10-1-1 will likely dictate the regulatory framework for the rest of the United States.