On May 6, the Consumer Financial Protection Bureau (CFPB or Bureau) released its Fair Lending Report for 2021. As in 2020’s report, published last year, the CFPB shows that its focus remains on what it characterizes as “financial inclusion, racial and economic equity, and fair competition”:
“As part of the prioritization process, the CFPB identifies emerging developments and trends by monitoring key consumer financial markets. If this field and market intelligence identifies fair lending risks in a particular market, that information is used to determine the type and extent of attention required to address those risks.”
The 2021 report makes several prominent mentions of the use of artificial intelligence (AI) and machine learning — a practice that Fair Lending Director Patrice Ficklin says that she is skeptical of, especially as a “cure-all for bias in credit underwriting and pricing.” The report reaffirms that the CFPB will be expanding its evaluation of AI and machine learning models as used by institutions, including in evaluating applicants for credit.
But while the focus on machine learning and AI is not new, there is something very interesting about the report. In it, the CFPB seems to be reframing and expanding the scope of what it considers “fair lending.” It does this by framing its actions against a prison financial services company and a nonbank immigration services provider as fair lending cases, rather than as garden-variety UDAAP (unfair, deceptive, and abusive acts and practices) cases, even though the allegations in both cases did not involve discrimination. We had previously understood that the CFPB would prioritize UDAAP and other nondiscrimination cases when it felt that the injured consumers were members of protected classes, and the inclusion of these cases in the report underscores that point. But if those cases were brought today, would there be allegations of “unfair” discrimination, consistent with the CFPB’s March 2022 updates to its UDAAP exam manual? We will have to wait and see.
Also interesting for their inclusion in the report are the October 2021 orders to large tech companies to provide information to the CFPB on their business practices. Specifically, the orders sought information on data harvesting and monetization, access restriction and user choice, and other consumer protections. It is unclear why these orders are included in the Fair Lending Report; the short narrative that mentions them says nothing about discrimination.
The CFPB’s 2021 Fair Lending Report confirms much of the messaging we have been hearing over the last year, especially the Bureau’s continued suspicion of machine learning and AI technologies. But it leaves unanswered the outstanding questions about the Bureau’s intentions with respect to the “discrimination as UDAAP” announcement made in March this year. We’ll continue to watch and report on those developments here.