On June 25, the CFPB issued its annual fair lending report covering its fair lending activity in 2023. The Bureau noted that in 2023 it undertook 28 fair lending examinations and announced four enforcement actions. It separately referred 18 matters to the Department of Justice.
In the supervisory context, the Bureau stated that two of the most frequently identified fair lending citations were related to the granting of pricing exceptions and HMDA violations. The CFPB stated that its supervisory examinations focused on the following three areas:
- Mortgage origination (including redlining, appraisal bias, and HMDA compliance). The CFPB’s mortgage origination examinations focused on redlining as well as assessing potential discrimination in underwriting and pricing processes, including weaknesses in fair lending-related compliance management. It also reviewed appraisal service providers to identify risks that may arise due to discrimination or bias. Finally, it conducted HMDA data integrity and validation reviews.
- Credit card marketing (including the use of alternative data). The CFPB reviewed companies’ digital marketing practices relating to credit cards, as well as lenders’ use of alternative data in credit card marketing.
- AI and machine learning in credit card originations. The Bureau’s credit card examinations included evaluation of AI and machine learning models used by credit card lenders in credit card originations, as well as assessing whether there were disparities in application, underwriting, and pricing processes, and whether companies searched for less discriminatory alternatives to the models they used.
Regarding enforcement matters, the CFPB highlighted two enforcement actions; one that it brought against a large national bank for illegally discriminating against credit card applicants, and the other against a Texas-based land developer and lender for allegedly selling Hispanic families flood-prone land without basic utility infrastructure, and then setting those families up with seller-financed loans they could not afford. The CFPB also highlighted HMDA-related enforcement actions against two separate “repeat offenders” for reporting incorrect HMDA data. (One which we discussed here.)
The CFPB also discussed its fair lending rulemaking and guidance, including its Section 1071 small business lending data collection rule, and the interagency proposed rule on the use of automated valuation models. The Bureau also touched on its circular on adverse action requirements that lenders must adhere to when using artificial intelligence models.
Putting it into Practice: Many the violations identified in the Bureau’s 2023 Fair Lending Report were similar to ones noted in the 2022 report. Moreover, the Bureau’s skepticism towards AI, both in underwriting and in appraisals seems to be a recurring theme. Financial institutions that use AI in any of their product or service offerings should take note of the Bureau’s focus and ensure their compliance efforts are current and their models are thoroughly tested.