CFPB Emphasizes Requirement to Disclose Reasoning Behind Adverse Credit Decisions
The CFPB reminded lenders that use complex decision-making algorithms that they must still provide specific and accurate explanations under the Equal Credit Opportunity Act ("ECOA") and in particular, Regulation B for denying credit or taking other adviser actions.
In a Consumer Financial Protection Circular published on May 26, 2022, the CFPB reminded companies that make credit decisions using "complex algorithms" predicated on artificial intelligence and/or machine learning that they cannot rely on the complexity of such algorithms to avoid satisfying their obligations under the ECOA and its Regulation B to provide a statement of specific reasons as to why an application for credit was denied and/or why an adverse action was taken. The Circular warns creditors and other loan decision makers that relying on the uninterpretable or "black-box" nature of complex algorithms to avoid providing an explanation to consumers will not excuse compliance with federal consumer protection laws; rather, to the extent such algorithms produce unknowable, adverse credit determinations, they will not be permitted for use in making credit decisions.
In a press release accompanying the Circular, CFPB Director Rohit Chopra stated that "[t]he law gives every applicant the right to a specific explanation if their application for credit was denied, and that right is not diminished simply because a company uses a complex algorithm that it doesn't understand."
Commentary
The Circular follows a growing regulatory priority at the local, state, federal and international levels to ensure that the use of artificial intelligence and machine learning in a variety of contexts does not result in unfair, biased, discriminatory or unexplainable outcomes adversely impacting consumers and employees. But rather than providing guidance on auditing standards or other methods at achieving explained and explainable results, the Circular and its accompanying press release focus on enforcement, portending an increase in CFPB scrutiny of automated decision-making by creditors going forward.