Colorado SB21-169 (unfair discrimination through data and algorithms)
A 2021 law, the first of its kind in the United States, prohibiting insurers from unfair discrimination resulting from external data and algorithms and requiring them to test their models, starting with life insurance.
Enacted in July 2021, Colorado's SB21-169 prohibits insurers from using external consumer data, algorithms or predictive models in a way that results in unfair discrimination based on race, color, national or ethnic origin, religion, sex, sexual orientation, disability or gender identity. Its reach lies in targeting effects rather than intent: a model that uses none of these characteristics can violate the law if proxy variables, zip code, credit history, purchasing behavior, produce disproportionately unfavorable treatment. The state's Division of Insurance implements the law line by line, starting with life insurance, with a governance regulation in force since November 2023 and quantitative testing requirements. Insurers must inventory their external data sources and models, document their governance and test for outcome disparities between groups. For actuarial science, the law sharply raises the question of the boundary between legitimate risk-based segmentation and indirect discrimination, which European law has approached another way.
Textbook case: a life insurer uses a score based on purchasing data to speed up underwriting without a medical exam. The tests required in Colorado by the 2021 law show that applicants from certain groups are routed to the longer, costlier medical exam twice as often as others with a comparable health profile: the insurer must fix the model or justify the gap by risk.
Follow the research
One letter a week. First, the week's underwriting note, when one is published: a dated, argued decision, signed by Paul Bernardski. Then what the observatory published on cyber risk, political risk and insurance. No account, no tracking.
No spam, one-click unsubscribe, no data ever sold.