One of the most important areas of consumer financial regulation today is the use of internet- or social media-based platforms to target advertising for consumer financial products. Its importance stems from the fact that financial services companies can obtain significant benefits from directing advertising to those consumers most likely to be interested in a product, and to be able to qualify for it. But the fair lending risks associated with the use of targeted marketing strategies have been manifested in a series of private lawsuits and a charge of discrimination filed by HUD against Facebook, with the concern being that protected characteristics, or close proxies, might be used to exclude consumers from seeing advertisements on bases that may violate the Fair Housing Act or Equal Credit Opportunity Act. Unfortunately, this is an area where there is very little regulatory guidance or commentary, leaving financial institutions with the task of making guesses and judgment calls about how to use ad targeting effectively, and in a way that minimizes fair lending risks.
Recently, Carol Evans and Westra Miller from the Federal Reserve’s Division of Consumer and Community Affairs published an article in the Fed’s Consumer Compliance Outlook addressing this very issue, entitled “From Catalogs to Clicks: The Fair Lending Implications of Targeted, Internet Marketing.” The article is helpful not only because it contains a thorough introduction to the manner in which targeted advertising works, but also a specific discussion of the fair lending risks posed by financial institutions’ use of targeted marketing, and guidance about how to avoid those risks. Because the article represents one of the few regulatory perspectives on this issue, we highly recommend that financial institutions read it.
The authors identify two primary risks from the use of targeted advertising: redlining and steering. Redlining, a legal theory traditionally associated with mortgage lending, is the more familiar theme of fair lending concern in this area, and stems from the possibility that consumers will be excluded from advertising for financial products because of prohibited factors like age, gender or race. Evans and Miller note that this carries with it the possibility of very significant consumer harm: “The growing prevalence of AI-based technologies and vast amounts of available consumer data raises the risk that technology could effectively turbocharge or automate bias. In doing so, we risk further entrenching past discrimination into future decision-making.”
The article also highlights steering concerns that may result from targeted marketing. The authors warn that curating products offered to a consumer carries with it the risk of steering consumers into less-favorable products than those they may qualify for, and if that steering runs along the lines of a protected characteristic, it could create a violation of fair lending laws. (Incidentally, in my view, even without the protected class correlation, such an approach to advertising could create a UDAP/UDAAP problem.)
The article makes specific recommendations to financial institutions considering the use of targeted marketing. It cautions that any use of the technology should be “approached with an awareness of the risks,” and be subject to fair lending compliance measures. With respect to the redlining risk, the article leaves no room for doubt that financial institutions need to be educated about the variables used in directing their advertising.
Lenders that use online advertising services or platforms can take steps to ensure that they monitor the terms used for any filters, as well as any reports they receive documenting the audience(s) that were reached by the advertising. It is also important to understand whether a platform employs algorithms — such as the ones HUD alleges in its charge against Facebook — that could result in advertisements being targeted based on prohibited characteristics or proxies for these characteristics, even if that is not what the lender intends.
Of course, understanding whether these characteristics are being used requires insight into the elements used in an ad-targeting model, which makes it difficult for a financial services company to safely use the ad-targeting feature of an internet or social media platform that refuses to share the elements used in its targeting model (many platform operations do not share the elements).
My perception, which is reinforced by Evans’ and Miller’s article – is that the fair lending risks involved here, and general principles of vendor management, will cause regulators to have little tolerance when a financial institution uses an ad-targeting platform without having adequate insight into the factors used to select consumers to receive advertisements. I believe that insight may be gained by a combination of direct review of the model variables, through concrete and specific representations of the ad-targeting platform, and by the financial institution’s own testing of applications/inquiries generated by an advertisement. But the bottom line is that a financial institution must do something to inform itself of how its advertisements are being placed.
With respect to the steering risk, the article makes a straightforward and intuitive recommendation – that “when a consumer applies for credit, she is offered the best terms she qualifies for, regardless of what marketing channel or platform was used to target marketing to the consumer or collect her application.”
Evans’ and Miller’s article underlines that the fair lending risks from targeted advertising are very real, and are being taken very seriously by financial regulators. We will continue to monitor developments in this area and bring them to you as they occur.