The CFPB issued a Circular on September 19, 2023 warning creditors using “artificial intelligence or complex credit models” that they may not rely on the checklist of reasons for adverse action in Regulation B. Rather, creditors must specifically and accurately indicate the “principal reason” for the adverse action. However, in our view the Circular raises more questions than it answers!
As most creditors know, the Equal Credit Opportunity Act (“ECOA”) and its implementing regulation, Regulation B, require creditors to provide an adverse action notice any time an applicant is denied credit or when an adverse action is taken with respect to an existing credit, such as a credit line decrease. For many years, Appendix C to Regulation B has included sample forms for the provision of adverse action notices. These forms include a section listing multiple potential “principal reasons” for the denial of credit. Creditors have viewed these principal reasons as a safe harbor for use in adverse action notices.
Unfortunately, instead of promulgating updated forms, the CFPB has fired a warning shot across the bow to creditors relying on sophisticated computer scoring models, including algorithms, artificial intelligence, and the use of data sources outside of traditional credit reporting. Using these types of scoring models, of course, it is often difficult for the creditor to pinpoint a specific principal reason that an applicant was denied credit. Rather, the reasons for credit denials are buried in the programming used by the scoring model.
While the Circular makes clear that creditors should not just check the “closest” option from the safe harbor list in the Regulation B templates, it doesn’t provide creditors with any real guidance as to what they should do instead if the model’s denial of credit isn’t capable of being distilled into a single discernible “principal reason.” Further, based on the language of Regulation B itself, a creditor may not simply note that an applicant “failed to achieve a qualifying score on the creditor’s credit scoring system.”
Given the CFPB’s Circular and the existing text of Regulation B, creditors should carefully evaluate their adverse action notices in any credit programs relying on complex credit scoring models. If the scoring model programming relies on the adverse action notice forms from Regulation B, it would be appropriate to dig deeper to determine if the reasons noted are truly the “principal reasons” for denials of credit. If not, the creditor should revise the adverse action notices to come into compliance by attempting to describe the principal reasons for denial more accurately. One option, of course, may be to ensure that the programming behind the scoring model generates a report noting the principal reason(s) for a denial.