A coalition of 14 state attorneys general recently issued joint guidance on antidiscrimination protections “to clarify the state of the law” for their residents and “reaffirm [their] commitment to defend and enforce civil rights laws.” The coalition included the attorneys general of California, Delaware, Hawaii, Illinois, Maryland, Massachusetts, Michigan, Minnesota, Nevada, New Mexico, New York, Oregon, Vermont and Washington.
The guidance indicates that state and federal civil rights laws support liability for both intentional discrimination and disparate impacts, notwithstanding recent federal actions aimed at narrowing or altogether eliminating the latter. It is an explicit response to the multiple efforts at the federal level to eliminate disparate impact liability – including by the Federal Trade Commission, Consumer Financial Protection Bureau (CFPB), Office of the Comptroller of the Currency and Department of Justice. Those efforts stemmed from a Trump administration executive order directing federal agencies to discontinue use of disparate impact liability when administering antidiscrimination statutes.
Overview of the guidance
In the guidance, the attorneys general reiterate their position that civil rights laws support two separate theories of liability: intentional discrimination, where a person or entity deliberately treats a protected group differently or denies it opportunities, and disparate impact, where a facially neutral policy or practice produces discriminatory effects on a protected group without adequate justification.
The attorneys general assert that recent executive orders, agency rules and agency memoranda characterizing disparate impact liability as unlawful in fact misstate the law and do not change the substantive protections available under the Constitution or state or federal statutes. Describing disparate impact liability as a “longstanding, constitutional tool,” they note the long history of federal courts affirming the disparate impact theory of liability under various statutes, including Title VII of the Civil Rights Act of 1964 and the Fair Housing Act. The attorneys general further note that while recent US Supreme Court decisions may have narrowed who may bring certain disparate impact claims, the decisions have not eliminated the underlying theory of liability.
The state attorneys general also highlight the importance of state and federal agency investigation and enforcement actions to root out and remedy antidiscrimination violations. As one example, they note that recent state investigations revealed certain biases embedded in AI and algorithmic decision-making systems used by certain lenders, landlords and employers that could have a discriminatory effect based on a protected characteristic, such as age, race or sex.
Recommended proactive measures for lenders
The state attorneys general conclude the guidance by encouraging businesses to adopt proactive measures to identify and address discriminatory effects to prevent potential harm. For participants in the housing and lending sectors, the guidance specifically recommends:
- Analyzing machine learning and algorithm-based lending and tenant screening tools throughout the tools’ life cycle to identify potentially discriminatory outcomes and refine the algorithms to include factors more relevant for evaluating creditworthiness.
- Routinely reviewing marketing strategies and lending and tenant screening policies and tools to identify potential bias and reduce barriers to housing.
The guidance frames these measures as consistent with existing law, cost-effective compared to after-the-fact enforcement, and beneficial to institutions seeking to manage fair lending risk proactively.
Looking ahead
The guidance does not alter the current federal regulatory landscape, including the CFPB’s recent amendments to Regulation B narrowing the availability of disparate-impact theories under the Equal Credit Opportunity Act. It does, however, confirm that a substantial bloc of states intends to continue applying disparate impact and intentional discrimination theories under their own civil rights and fair lending statutes, independent of federal policy shifts.
Financial institutions operating in any of the 14 states that joined this guidance should consider reviewing underwriting models, marketing practices and algorithmic decision-making tools used in lending and tenant screening against both federal and applicable state fair lending standards, with particular attention to effects-based outcomes rather than facial neutrality alone.