Prohibits the use of external consumer data and information sources being used when determining insurance rates; provides that no insurer shall unfairly discriminate based on race, color, national or ethnic origin, religion, sex, sexual orientation, disability, gender identity, or gender expression; or use any external consumer data and information sources, as well as any algorithms or predictive models that use external consumer data and information sources, in a way that unfairly discriminates based on race, color, national or ethnic origin, religion, sex, sexual orientation, disability, gender identity, or gender expression; makes related provisions; defines terms.
A04427 would amend New York’s Insurance Law to restrict insurers’ use of external consumer data and information sources, including algorithms and predictive models that rely on such data, when setting rates or otherwise conducting insurance practices. The bill prohibits unfair discrimination based on race, color, national or ethnic origin, religion, sex, sexual orientation, disability, gender identity, or gender expression, and directs the Superintendent of Financial Services to adopt rules to implement these standards.
The bill establishes a regulatory framework requiring insurers to disclose to the superintendent what external data sources they use, explain how those sources are incorporated into underwriting, pricing, and related practices, and maintain risk-management processes to test for discriminatory effects. It also requires insurer attestations, ongoing monitoring, and a remediation period if an algorithm or model is found to have an unfairly discriminatory impact. The superintendent would be authorized to examine and investigate insurer practices, issue annual reports to state leaders, and rely on stakeholder meetings with carriers, producers, consumer advocates, and other interested parties before finalizing rules.
The bill would add a new section 2403-a to the Insurance Law, creating statewide standards governing the use of external consumer data, algorithms, and predictive models in insurance practices. It would affect insurers across most lines of insurance, while expressly excluding title insurance, surety bonds, and most commercial insurance policies, with a limited exception for certain small business owners’ policies and commercial general liability policies with annual premiums of $10,000 or less. It also preserves the use of traditional underwriting factors in life, annuity, long-term care, and disability insurance, subject to the bill’s limits when those factors are part of a broader model using external consumer data.
The measure would give the Department of Financial Services new oversight and enforcement responsibilities, require rulemaking, and treat insurer-submitted materials as confidential trade secrets. It would likely increase compliance obligations for insurers that use alternative data, machine learning, or other predictive tools in pricing and underwriting, while creating a formal process for reviewing whether those tools produce discriminatory outcomes.
The available context suggests the bill is framed as a consumer-protection and civil-rights measure aimed at preventing bias in insurance pricing and underwriting. Its caption and text indicate a strong policy preference for transparency, testing, and oversight of data-driven insurance models, especially where they may disadvantage protected classes. No committee transcript or vote record is provided, so there is no documented recorded debate or roll-call sentiment in the supplied materials.
The main points of contention are likely to center on the scope of the prohibition and the compliance burden on insurers. Consumer advocates would likely support the bill’s anti-discrimination safeguards and transparency requirements, while insurers may object to limits on proprietary models, the cost of testing and reporting, and uncertainty over what counts as “external consumer data” or “unfair discrimination.” Another likely issue is the bill’s treatment of algorithms that correlate with protected characteristics but are used for actuarial purposes, as well as the confidentiality provisions that shield submitted materials from public disclosure.