On September 19, 2023, the Consumer Financial Protection Bureau (CFPB) issued Circular 2023-03 to provide creditors additional guidance for communicating denials of credit to consumers based upon artificial intelligence (AI) or complex credit models. The circular makes clear that it is insufficient for a creditor utilizing such models to satisfy their obligation under the Equal Credit Opportunity Act (ECOA) and Regulation B by relying on codified sample adverse action notice forms if the forms do not accurately indicate the principal reason for the denial. The circular also provides that creditors may not rely on overly broad statements of denial to the extent that such statements obscure the specific and accurate reasons for the credit decision.

Adverse action notification requirements under ECOA and Regulation B

Regulation B requires a creditor that denies an application for credit to provide to the applicant an adverse action notice containing a statement of reasons for the action. The statement must be “specific” and indicate the “principal reason(s) for the action taken.”¹ The reasons disclosed to the applicant must “relate to and accurately describe the factors actually considered or scored by a creditor.”² Regulation B contains sample adverse action notification forms with an illustrative – but not exhaustive – list of reasons for taking adverse action, and the CFPB states that a creditor may use the checklist of reasons in the sample forms only to the extent that the reasons listed “are specific and indicate the principal reason(s) for the adverse action taken.”

The circular highlights the provision of Regulation B that states: “If the reasons listed on the forms are not the factors actually used, a creditor will not satisfy the notice requirement by simply checking the closest identifiable factor listed.” Rather, the sample forms merely provide an illustrative and nonexclusive list. Thus, if the principal reason(s) a creditor actually relies on is not accurately reflected in the checklist of reasons in the sample forms, it is the duty of the creditor – if it chooses to use the sample forms – to either modify the form or check “other” and include the appropriate explanation, so that the applicant against whom adverse action is taken receives a statement of reasons that is specific and indicates the principal reason(s) for the action taken.

Adverse action notification complications associated with denials based on use of AI and complex credit models

The CFPB’s circular notes the heightened import of – and additional complexity associated with – providing specific reasons for adverse actions when creditors use AI in making credit determinations, because consumers may not anticipate that data gathered outside of their credit application or file may be the principal reason for the adverse action taken against their credit application. As one example, the circular states: “[I]f a complex algorithm results in a denial of a credit application due to an applicant’s chosen profession, a statement that the applicant had ‘insufficient projected income’ or ‘income insufficient for amount of credit requested’ would likely fail to meet the creditor’s legal obligations,” even if the indicated reasons are broadly related to the actual reason for credit denial.

Along those lines, the CFPB also indicates that, where a creditor lowers a consumer’s credit limit based on behavioral data, such as where the consumer shops, an adverse action reason stating “purchasing history” or “disfavored business patronage” would be insufficiently specific. To satisfy Regulation B, the CFPB states the creditor would need to disclose additional specific details, such as the type of establishment, location of business or type of goods purchased, that led to the adverse action.

Other recent CFPB actions related to AI/fair lending

This AI-related fair lending circular builds upon the CFPB’s Circular 2022-03, published last year, which stated that the ECOA requires creditors to explain the specific reasons for taking adverse actions against consumers. The CFPB also last year issued an advisory opinion reiterating that creditors must provide adverse action notices to borrowers when changes are made to existing lines of credit – and not only for applications for new credit. The series of issuances underscores the CFPB’s general policy concerns regarding the effects of technology on fair lending compliance. For example, the CFPB issued guidance last year informing corporate landlords that prospective tenants must receive adverse action notices when denied housing. The CFPB also participated in an interagency rulemaking regarding automated valuation models designed to ensure that AI models do not exacerbate disparities based on historical patterns of discrimination.

Looking ahead

In its most recent annual report, the CFPB noted it is increasing its expertise in data science and fair lending analytics to ensure it is able to hold creditors and service providers accountable for complying with fair lending laws. In addition to conducting general fair lending testing of advanced models, this new CFPB circular makes clear that creditors also will need to carefully consider how they communicate to consumers credit decisions made on the basis of AI and complex credit models.

Notes

1. 15 USC 1691(d)(3); 12 CFR 1002.9(b)(2).
2. 12 CFR Part 1002 (Supp. I), sec. 1002.9, para. 9(b)(2)-2.

[View source.]