CFPB: Credit Denials Must Have Specific, Accurate Explanations

credit application

The Consumer Financial Protection Bureau (CFPB) wants to remind consumers that federal law requires companies to explain specific reasons for denying credit applications, even when the creditor employs credit models using complex algorithms, according to a Thursday (May 26) press release.

Citing the Equal Credit Opportunity Act (ECOA), the agency said the law applies regardless of whether a firm is using a “black-box model” to make decisions on credit applications.

“Companies are not absolved of their legal responsibilities when they let a black-box model make lending decisions,” CFPB Director Rohit Chopra said in a news release. “The 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.”

See also: CFPB’s Opinion on Fair Lending Rules Could Extend to AI

ECOA protects applicants against discrimination when seeking, applying for, and using credit by making sure creditors provide a notice when taking adverse action against an applicant. In that notice, the creditor needs to list specific, accurate reasons for their action. They cannot use technologies in decision-making if that technology keeps them from being able to provide these necessary explanations, the bureau said.

The CFPB notes that law-abiding financial companies have for years used advanced computational methods to make credit decisions, and have been able to show the thinking behind those decisions.

“However, some creditors may make credit decisions based on the outputs from complex algorithms, sometimes called “black-box” models,” the bureau said. With such models, adverse action notices that meet ECOA’s requirements may not be possible.”

Earlier this month, the CFPB said released its annual Fair Lending Report to Congress, which looks to the future of the CFPB’s work with financial markets and mentions predictive analytics, algorithms, and machine learning. It says that the agency is worried that “while the technology holds great promise, it can also reinforce historical biases.”

Read more: CFPB Targets Credit Card “Suppressed Data” Practices

This week also saw the CFPB release research that found many of the nation’s largest credit card companies aren’t routinely providing data to credit bureaus on the actual monthly payments their borrowers are making.

The bureau said its research found that only half of these companies contribute data to credit reporting companies about the exact monthly payment made by borrowers and that many of them suppressed information they had in the past furnished.