New York Department of Financial Services Issues Guidance Regarding Life Insurers’ Use of External Consumer Data in Underwriting
On January 18, 2019, the New York State Department of Financial Services (NYDFS) issued Circular Letter 2019-1 (the Circular Letter), addressing insurers’ use of external consumer data and information sources in underwriting for life insurance. The Circular Letter follows an investigation commenced by NYDFS regarding life insurers’ use of external data, which was initiated in light of reports that insurers were using algorithms and predictive models that include unconventional sources or types of external data. Among other things, the Circular Letter provides guidance that when insurers use external data sources in connection with underwriting decisions, (1) the use of external data sources must not result in any unlawful discrimination, (2) the underwriting or rating guidelines must be based on sound actuarial principle; and (3) life insurers must have adequate consumer disclosures to notify insureds or potential insureds of the right to receive the specific reasons for any adverse underwriting decision based on such data.
Based on its investigation, NYDFS determined that life insurers’ use of external data sources in underwriting has a strong potential to mask prohibited discriminatory practices and that use of certain algorithms and predictive models may also lack a sufficient rationale or actuarial basis. The Circular Letter defines “external data” to include any data or information sources not directly related to the medical condition of the applicant that is used, in whole or in part, to supplement traditional medical underwriting, as a proxy for traditional medical underwriting or to establish “lifestyle indicators” that may contribute to an underwriting assessment of an applicant for life insurance coverage. The Circular Letter reiterates that existing laws prohibit life insurers from making underwriting decisions based on race, color, creed, national origin, status as a victim of domestic violence, past lawful travel or sexual orientation and limit insurers from relying on physical or mental disability, impairment or disease, or history of disability or disease. Importantly, NYDFS notes that insurers are responsible for complying with the anti-discrimination laws regardless of whether an insurer itself collected the data or whether the insurer is relying on external data sources, algorithms of external vendors or predictive models that collect or use prohibited data. In other words, insurers may not use an external data source to collect or use information that the insurer would otherwise be prohibited from collecting or using directly. The Circular Letter clarifies that the burden of compliance with such anti-discrimination laws rests with the insurer and that diligence will be required when insurers receive external data from third-party vendors.
In addition, when evaluating whether an underwriting or rating guideline derived from external data sources is unfairly discriminatory, insurers should consider (i) whether the underwriting or rating guideline is supported by generally accepted actuarial principles or actual or reasonably anticipated experience that justifies different results for otherwise similarly situated applicant and (ii) whether there is a valid explanation for the differential treatment of similarly situated applicants. This includes for example, models and algorithms that purport to make predictions about a consumer’s health status based on factors such as the consumer’s retail purchase history, social media, internet or mobile activity, geographic location tracking, the condition or type of the consumer’s electronic device or the consumer’s appearance in a photograph. The Circular Letter specifically notes that even if statistical data supports an underwriting or rating guideline, there must still be a valid rationale or explanation supporting the differential treatment of otherwise similar risks. Therefore, an insurer may not rely on external data or external predictive algorithms or models unless the insurer has determined that the external data or predictive model is otherwise permitted by law or regulation and is based on sound actuarial principles or experience that justifies any resulting differential treatment of otherwise like risks.
The Circular Letter also clarifies that when life insurers use external data sources in making underwriting decisions, they must adhere to Section 4224(a)(2) of the New York Insurance Law and notify the insured or potential insured of their right to receive the specific reasons for a declination, limitation, rate differential or other adverse underwriting decision, including, controversially, the failure to qualify for an accelerated underwriting process as compared to traditional medical underwriting. This includes the specific source of the information on which the insurer based its adverse underwriting decision. The insurer may not rely on the proprietary nature of a third-party vendor’s algorithmic processes to justify a lack of specificity. The Circular Letter goes on to say that failure to adequately disclose to a consumer the material elements of an accelerated or algorithmic underwriting process, and the external data sources on which it relies, may constitute an unfair trade practice.
Life insurers interested in using external data sources, algorithms or predictive modeling in accelerated underwriting processes should exercise caution to ensure that the use of data contained in such materials is not unlawfully discriminatory and would otherwise be permitted by law or regulation. In addition, life insurers should be aware that they are responsible for establishing that the external data sources, algorithms or predictive models are based on sound actuarial principles and that they must notify insureds or potential insureds of their right to receive the specific reasons for any adverse underwriting decision. Finally, life insurers are ultimately responsible for ensuring compliance with such laws through diligence, even in the event that such external data, algorithm or predictive model is provided by a third-party vendor.