Colorado Bill Would Restrict Insurers’ Use of Consumer Data and Algorithms

The Colorado General Assembly is considering a bill that would restrict insurers' use of external consumer data, algorithms and predictive models that have “no direct causal relationship” to the risk being insured and that result in unfair discrimination based on an individual's membership in a protected class. The bill would prohibit certain practices and give the Colorado Division of Insurance greater insight into how insurers are using data and predictive models.

Among other things, SB21-169 would:

• Prohibit insurers from considering an individual’s race, color, national or ethnic origin, religion, sex, sexual orientation, disability or transgender status in any insurance practice. “Insurance practice” is defined broadly and includes marketing, underwriting, pricing, utilization management, reimbursement methodologies, claims settlement and fraud detection.
• Prohibit insurers from directly or indirectly using any external consumer data and information source, algorithm or predictive model that unfairly discriminates against an individual based on the newly defined prohibited factors. “External consumer data and information source” is defined broadly and includes credit scores, social media habits, locations, purchasing habits, home ownership, education, licensure, civil judgments and court records. The bill would not create a private right of action.
• Require an insurer that uses external consumer data and information sources, algorithms or predictive models to submit certain information to the Colorado Division of Insurance in a form, manner and frequency to be determined by the Commissioner. The submission would:
  • Describe the external consumer data and information sources used by the insurer in the development and implementation of algorithms and predictive models.
  • Indicate each insurance practice in which the insurer is using the data, algorithms and models.
  • Explain the manner in which the insurer is using the data, algorithms and models.
  • Include an attestation that the data, algorithms and models used by the insurer do not (i) intentionally or unintentionally use information concerning the prohibited factors, or (ii) result in proxy discrimination based on the prohibited factors. (Like the NAIC’s Principles on Artificial Intelligence, the bill does not define proxy discrimination.)
  • Include an assessment of whether the insurer’s use of the data, algorithms and models may result in unfair discrimination based on the prohibited factors and, if so, describe what actions the insurer has taken to minimize the risk of unfair discrimination, including ongoing monitoring.
• Authorize the Commissioner to examine and investigate an insurer’s use of external consumer data and information sources, algorithms or predictive models. If the Commissioner determines that the use of an external consumer data and information source, algorithm or predictive model (i) bears no direct causal relationship to insurance losses or to the condition of the property or applicant to be insured, and (ii) unfairly discriminates based on a prohibited factor, the Commissioner may promulgate rules restricting or prohibiting the use of the information, algorithm or predictive model.

The Colorado bill is the latest example of how the use of external consumer data, algorithms and predictive models may subject insurers to greater scrutiny. If your organization is using these tools, it’s time to consider how your systems measure up to the regulators’ expectations. Our lawyers and data scientists stand ready to assist in that effort.