In their June 2021 request for information regarding financial institutions’ use of artificial intelligence (AI), including machine learning, the CFPB and federal banking regulators flagged fair lending concerns as one of the risks arising from the growing use of AI by financial institutions.
Last week, in an apparent effort to increase its scrutiny of machine learning models and those that use alternative data, the CFPB published a blog post titled “CFPB Calls Tech Workers to Action,” in which it made a direct appeal to “engineers, data scientists and others who have detailed knowledge of the algorithms and technologies used by companies and who know of potential discrimination or other misconduct within the CFPB’s authority to report it to us.”
In the blog post, the CFPB describes whistleblowing as “a tool to hold industry accountable” and observes that “[t]ech workers may have entered the field to change the world for the better, but then discover their work being misused or abused for unlawful ends.” It states that AI “can help intentional and unintentional discrimination burrow into our decision-making systems, and whistleblowers can help ensure that these technologies are applied in law-abiding ways.” The following scenario is provided as an example:
[W]hile algorithmic mortgage underwriting is sometimes hailed as a method to significantly reduce housing discrimination, and many of those designing the algorithms seek to create a fairer housing market, that’s not always how things work out. In a recent study of over 2 million mortgage applicants, researchers found discriminatory effects of these new technologies, as Black and Hispanic families have been more likely to be denied a mortgage compared to similarly situated white families.
One researcher described the situation as one where loan officers take applicant information, but algorithms make the decisions. Whether such a process removes or embeds discrimination depends on a number of factors, including the types of data collected, how they are weighted, and how decisions are reviewed.
Dodd-Frank Section 1057 protects whistleblowers who report alleged violations of the Consumer Financial Protection Act, any law subject to the jurisdiction of the CFPB, or any CFPB rule, from retaliation by their employers. These protections apply if the employee reports the alleged violations to the employer, the CFPB, or any other federal, state, or local government authority or law enforcement agency. However, Dodd-Frank does not currently provide financial incentives for whistleblowers who report such alleged violations. (In 2020, the Bureau proposed legislative language to amend Dodd-Frank to establish a program for whistleblowers to receive monetary awards.)
Given the lack of financial incentives for whistleblowers, we doubt that this direct appeal will bear much fruit for the Bureau. It does, however, underline the seriousness with which the Bureau has taken up criticisms of newer-technology underwriting methods, and highlights that financial institutions must be in a position to defend their models when they are inevitably scrutinized.