Targeted Ads, Fair Lending Risk: Why Lenders Should Plan for Today’s and Tomorrow’s CFPB

Podcast: The Consumer Finance Podcast
Episode: Targeted Ads, Fair Lending Risk: Why Lenders Should Plan for Today’s and Tomorrow’s CFPB
Host: Chris Willis
Guest: Stephen Hayes
Aired: October 29, 2026

Chris Willis (00:05):
Welcome to The Consumer Finance Podcast. I’m Chris Willis, the co-leader of Troutman Pepper Locke’s Consumer Financial Services Regulatory Practice, and today we’re joined by a special guest and friend to talk about one of those mystifying topics in credit, and that is the fair lending implications of targeted advertising, both today and how it may look in the future. But before we jump into that very interesting topic, let me remind you to visit and subscribe to our blogs, which now have all been unified under a single financial services hub. That’s right, all of our blog content, all of our podcasts, and everything is all in one easy-to-get place, and you can find it at troutmanfinancialservices.com. And speaking of our other podcasts, don’t forget to give them a listen. We’ve got; The FCRA Focus, The Crypto Exchange, Payments Pros, and Moving the Metal. Those are all available not only through our hub, but also on all popular podcast platforms. And speaking of those platforms, if you like this podcast, let us know. Leave us a review on your platform of choice and let us know how we’re doing.

Now, as I said, today we’re going to be diving into a topic that has been mystifying, at least for me and probably for a lot of other people for a long time, and that is, what are the fair lending implications or rules around targeted online advertising? And joining me today to talk about that important topic is a previous guest and friend, Stephen Hayes. Stephen is the co-managing partner of the Relman Colfax firm, has been a fair lending expert for many, many years, and I really appreciate him being on the podcast today. So Stephen, thanks for being here.

Stephen Hayes (01:34):
Thank you. I’m excited to be here.

Chris Willis (01:35):
So let me just try to set the stage here, because the idea of targeted advertising has been kicking around for, I feel like, six or seven years now with some cases brought by consumer advocacy groups against social media platforms. The National Fair Housing Alliance brought one, as I recall, in around 2018 or so that was settled. There was a Washington Attorney General settlement with the same social media platform. And then in 2019, HUD issued a charge of discrimination and an investigation under the Fair Housing Act against the same social media platform for targeting of ads for housing and housing-related services. And that investigation went on for a number of years and resulted in a DOJ lawsuit being filed against the social media platform in the summer of 2022, which was then settled in the early part of 2023. But beyond that, there hasn’t been a lot of, to my knowledge, regulatory guidance or discussion about targeted advertising. And a lot of what we have so far has centered around housing. Both the private litigation and the HUD and DOJ proceedings were brought under the Fair Housing Act. And so it leaves providers of non-mortgage credit, I think, a little bit confused. And so I really wanted to get your perspective and let our audience hear it about where are we on this issue today and where might we be on it under a future administration where perhaps there’s a Democrat in the White House?

Stephen Hayes (03:04):
Yeah, I agree. It’s an area where we have not really gotten, I think, sufficient guidance. It is a hugely important area. So I remember, and I think you and I probably talked about it when it came out, but back in 2019, Carol Evans wrote for the Federal Reserve Board’s Consumer Compliance Outlook a piece talking about the fair lending risks raised by targeted digital marketing. And I think that that’s still a great article to read in terms of summarizing the types of steering risks, the types of disparity risks, the disparate treatment, disparate impact risk, all of those. I think those risks have gotten, in a sense, actually more complicated and more acute. And part of that is, at least I’m seeing, a relatively high level of sophistication in digital marketing, and sometimes that marketing starting to bleed into areas that feel like underwriting or pricing-adjacent areas. And so, actually, this, I remember Jason from your firm years ago wrote an article talking about how increasingly some marketing tools can predict not just who’s going to be interested in a product, which is what you would expect and can be extremely efficient at doing that, but also who would qualify for that product. And so you’re starting to integrate information in that feels a bit like underwriting information.

I have also seen to some extent pricing too. So pricing, depending on what channel, what type of ad people respond to, how are they interacting with a lender, can affect their pricing. And sometimes that can be not because risk is viewed differently among those populations, but there’s a bit of a willingness to pay because of the advertising channel that they come in on. And so we start to see it being quite good at picking up who might be interested in a product, but also starting to bleed into pricing and underwriting in ways that I think actually create a pretty core tension. Now the last piece I’ll say, and then I suspect we’ll talk a lot more about this, is we started to see the regulators in the last administration get focused on this. HUD issued some guidance about the Fair Housing Act. The CFPB, I think, was interested in this issue but didn’t take a lot of public action related to it. I think that’s going to change. I think people are realizing, and certainly the next wave of regulators are going to realize that advertising is not just the funnel for who comes into your financial institution, but it’s also going to dictate pricing and terms in ways that we haven’t seen before. And so it’s not going to be a choice. They’re going to have to focus on this.

Chris Willis (05:25):
Yeah, let me just follow up on a couple of things you mentioned. First of all, I agree Carol’s article is very important. She then subsequently went to the fair lending office of the CFPB before retiring. But for anybody who wants to find that article, as you noted, it’s in the Fed’s Consumer Compliance Quarterly and it’s called something like, “From Catalogs to Clicks.” And Carol Evans is the first author on it. So it should be easy to find on the Federal Reserve’s website. And then let me agree with you about the sort of current landscape of how industry is using targeted marketing, because I regularly encounter advertising models that have components that are designed to predict not only interest in or need for the product, but also likelihood of qualification. Because again, from an advertising standpoint, you don’t want to advertise the product and get somebody to apply who’s going to get declined. It’s not a good customer experience, and it’s also not a good use of your advertising money. But there have been some instances where an advertising campaign is sort of the gateway through which someone has to go to get special terms or more favorable underwriting consideration or something like that. And that, I think, creates an even greater overlap between the advertising and the underwriting decision. Because if that ad is the gateway to either the product or special terms under the product, in my view, we have to look at it differently than purely an advertisement. At least that’s my view. So as the industry has become more sophisticated about this, I agree the landscape presented to a regulator who gets interested in this in the future will be very different even than what Carol was talking about when she wrote that article, which was seven years ago.

Stephen Hayes (07:01):
We can talk a little bit about, I think, the landscape for the platforms has changed too. If we go back to the National Fair Housing Alliance lawsuit that you mentioned, the allegations there were that the platform contained pre-populated lists that allowed advertisers for housing, employment, and credit to actually exclude protected classes. So you could exclude African-Americans, you could exclude Hispanics, you could exclude Asians. You couldn’t exclude Whites, if I recall correctly. But it could not be more explicitly tied to protected class than was true then. That platform changed. They created a separate portal for housing, employment, and credit ads. Other platforms also created special portals to say, well, let’s limit that type of explicit targeting. We saw that shift that I think matters. That platform also, as a result of the DOJ settlement, implemented what it called a variance reduction system that essentially seems to mitigate disparate impact type risks. I believe that that DOJ settlement now has run its course, and it has concluded as of June. So will that continue? I don’t know. But I will say that we’re a little bit at a pivot point in my mind, which is some of the platforms have reacted to some of this. They’ve made certain changes. Are those sufficient? I don’t know. Certainly financial institutions have become very careful about a lot of this. But we’re also at a time where the CFPB at least has really pulled back from this space, not just in terms of prioritizing it, but actually in terms of changing Reg B so that Reg B literally would not cover a lot of the conduct that were frankly part of the allegations against Facebook, even in 2018 or whatever that lawsuit was. So I think that’s a pretty significant shift. We’re going to see the markets either really leaning in to using pretty discriminatory criteria or holding the line, understanding that there might be future risks coming in the next administration.

Chris Willis (08:55):
Let’s talk about the conduct that we’ve been warned off of by the regulatory action on this issue so far. And when I read the allegations in the National Fair Housing Alliance lawsuit, which is just a private lawsuit, but the HUD charge from 2019, the Washington Attorney General settlement, the DOJ complaint that was filed in the summer of 2022, they all seemed to revolve around the explicit use of protected characteristics in directing advertising. And you gave that example too, where the advertiser could say, I want to exclude people, and it was protected classes. And so I felt like it was sort of obvious to everybody after all that action occurred that we should not target advertising based on explicit use of protected characteristics. And the HUD charge also talked about close proxies, like language preference or stuff like that that’s very closely related to a protected characteristic. So I feel like I absorbed that lesson from the regulatory activity, and I think the industry did too. And so I think everybody knows not to do that now. The open question, though, in my mind was after all that was to what extent, if at all, is there room for disparate impact analysis in advertising, either under the Fair Housing Act or under ECOA? And you referenced the HUD guidance that came out in the spring of 2024, and there HUD explicitly took the position that an advertising model is subject to disparate impact. It mandated disparate impact testing and less discriminatory alternatives for those advertising models for housing. That’s the only regulator I’ve heard say that out loud. And so it leaves a question in my mind, like to what extent should we think about disparate impact as opposed to disparate treatment, which I think everybody knows you shouldn’t do?

Stephen Hayes (10:41):
I would separate this out a little bit. One of the questions is, well, as a legal matter, could someone plausibly allege disparate impact violation of marketing? I think there’d be a very strong case for that under the Fair Housing Act. I also think that there are some state laws that clearly cover marketing and either explicitly allow for disparate impact or could quite likely be interpreted to allow for it. So I think that that state law risk is real. And I think that the ECOA risk is real too. We can talk about some of the rule changes that the CFPB has made and went into effect in July, but for a number of reasons, I think that those are not very durable changes. And so I think actually there’s quite a strong argument even for ECOA. And then the last piece is actually for the most explicit cases are the type of dropdown menu, I want to exclude Black people from this audience. And I think that those are quite explicit. But there’s for a long time, and even going back to Carol’s article and before that, there have been a lot of people warning about other dangers, like, for example, using a seed audience to create a custom audience that looks like your own population. And if your own population is predominantly White and you use that seed audience, then the algorithmic tool is very likely to try to identify a predominantly White audience to mirror your own. And whether you phrase that as a disparate impact risk because you’re not thinking, you’re not aware of that, and the neutral practice is you’ve used your own audience, I think you could also easily make a disparate treatment claim there. We are all aware of these risks. We’ve all talked about them for a long time. Any institution should be aware of it. And I think that some of this disparate impact, disparate treatment can actually bleed into one another. And so something that could be characterized as disparate impact could well be characterized as disparate treatment too, especially because we’re so aware of these risks. So I would keep that in mind too.

Chris Willis (12:30):
And I would just underline the last statement that you made that there’s bleedover between treatment and impact. I think that is a statement that is true in this area but is also true of many other fair lending areas.

Stephen Hayes (12:42):
Absolutely.

Chris Willis (12:43):
And if we were to envision a world where the U.S. Supreme Court were to say there’s no disparate impact under ECOA, the statute doesn’t allow for it, which they definitely could do, the regulators in the future, if they get motivated on fair lending issues as they have been in the past, I think have the option to characterize claims that they might have said were impact in the past as treatment in the future. And so the idea that even a Supreme Court decision would wash disparate impact out of our hair altogether, I think is not an accurate one.

Stephen Hayes (13:16):
I fully agree.

Chris Willis (13:17):
I want to talk a little bit more about how disparate impact might work here before we talk about the Reg B rule changes, which we definitely will get to in just a second. But let’s say I’m an advertiser and I have a model that uses a variety of information, including some stuff about creditworthiness, income, stuff like that, underwriting-type attributes. And my business justification for that is I only want to show the ad to people who are interested in the product and who are likely to qualify if they apply. To me, that feels like a business justification. How do you think that would play out if that strategy results in a disparate impact in the advertising? And how would we measure that disparate impact? Would we measure it against the American population as a whole? Or is it valid to limit the comparison population to those who actually would be interested in and qualify for the product? How would that work? Because I don’t think anybody’s actually done that that I’ve seen, either in supervision or in an enforcement case.

Stephen Hayes (14:13):
There’s not a single methodology to say, well, this is how you do it. I think that’s quite context-specific depending on are you talking about a campaign that you yourself, the institution, is running and you’ve got complete control over. In that case, that is actually not an uncommon thing for a lot of financial institutions to do, to do a disparate impact analysis of their individual marketing campaigns. They estimate who they think the likely audience is that would receive it. As you say, they gut check, well, what are the segmentations that we’re using that are driving those changes? And if there are differences that are meaningful, making sure that they’re both justified and then also exploring whether is there a way to still meet those business goals and not have as significant an impact. And so that’s actually, certainly there’s many ways to do it, but it’s not an unsolvable problem. Institutions do that on their own campaigns frequently. If you’re dealing with a third-party platform, it can be a lot more difficult. And those platforms, at least in my experience, vary quite a bit in terms of how much information they’re willing to give. But I’ve seen certainly custom third-party models where they are willing to do a similar exercise and tweak the model to make sure that they are eliminating disparities or at least mitigating those disparities.

For some other advertisements, I mean, the pool of who is the correct denominator in terms of who’s seeing it, who should be seeing it, I think is a difficult question, but not an unanswerable question at all. And I think there’s several different reasonable ways to do it depending on what the campaign looks like and who you’re trying to reach. But that’s frankly true with many disparate impact and, frankly, disparate treatment analysis too. There can be difficult problems about estimations and what the right comparator group is. And you and I have both been involved in battle of the experts about what the right approach is. But I don’t think it’s a problem where we say, well, there’s no possible way to do it. I think obviously there are lots of very reasonable ways to do it.

Chris Willis (15:59):
Let’s shift the conversation to something you alluded to a minute ago, which is the CFPB finalized in April of 2026 a revision to Reg B, which then became effective in July of 2026, which made significant limitations to the concept of discouragement under Reg B. Discouragement’s been in Reg B, I think, since the ’70s, and these restrictions are new. And part of it is this sort of higher burden of proof that the alleged person who’s discouraging had to know that it would discourage people on the basis of a protected characteristic. And there are also these statements of, targeting one group to receive an ad does not equal discouragement of another group. There’s that kind of commentary associated with it. So let’s talk about what that would do today. And you mentioned that you thought those changes might be temporary. I’d like to hear your views on both the present state of play and what the future may hold with respect to these issues.

Stephen Hayes (16:55):
I will note that changes made to Reg B that became effective in July, there’s been a lot of attention on the disparate impact changes and the special purpose credit program changes. I’ve heard a little bit less public focus on the discouragement changes, but in my mind they are just as significant. As you said, for decades, Reg B has made it illegal to discourage prospective applicants on the basis of a protected class. And it’s really obvious why, because that is a very easy way to evade ECOA. If you want to avoid lending to certain consumers, you just make sure that people from that group never apply in the first place. And it’s not rocket science that that would evade ECOA’s primary requirements. One of the changes that they made in Reg B that you didn’t mention is I think maybe one of the most problematic, which is that they say Reg B used to prohibit discriminatory acts or practices that would discourage receipt of applications. Now it says it only prohibits statements that would discourage someone. And so if we go back and think about the type of targeting that the National Fair Housing Alliance alleged Facebook was doing, saying, “Here’s a drop-down menu. Tell us who you don’t want these ads to go to. If you don’t want them to go to women, that’s fine. Click women. If you don’t want them to go to older consumers, that’s fine. Exclude them.” That would no longer be prohibited under the Reg B changes because it’s hard to characterize that type of exclusion decision as a statement. It is certainly an act or practice that would discourage getting receipt of those applications. And so that’s a pretty remarkable change.

Essentially what the CFPB is saying is ECOA has nothing to say at all about any of these marketing decisions or acts or practices. And so it would actually permit what you and I think both agree are some of the worst, most blatant discriminatory acts, I guess, whether you think they’re a violation or not. So that to me is really critical. Now, why is that not durable? I think there’s a few reasons. One, we are representing the National Fair Housing Alliance and others challenging that rule on APA grounds, so there’s ongoing litigation challenging it. And there’s also, even putting aside the possibility that the rule gets vacated through that litigation, it’s very easy here for a future administration to come in and just change the rule. And they don’t have to change it back to what it was before. They could change it so that it’s significantly more aggressive than what it was before. So you and I talked a few years ago when the Seventh Circuit heard a case called Townstone. And in that case, the Seventh Circuit said, “Well, yeah, of course, ECOA prohibits discouraging prospective applicants, and the CFPB has a lot of authority to prevent evasions of ECOA.” And so I think it’s quite likely that a future CFPB looks at Townstone and looks at the importance of marketing and says, “You know what? We’re going to change the rule so that it covers all types of marketing conduct, and there’s no question about our ability to go in and supervise and enforce there.” And so I would really keep that in mind. Both of those, I think, are really significant risks.

Then the last piece on marketing is we think about ECOA’s coverage, but oftentimes marketing is used not as an independent claim, but as circumstantial evidence to support some other claim. So if we think about redlining, which has always been and is illegal now, the practice sometimes there might be denial disparities, there might be not serving an area, and marketing might be used as evidence of intentional discrimination. You intended to not serve these customers. So it could always be looped in that way. So for all of this, I think it’s a remarkable change that the CFPB made. And it’s, to me, at a time when they should be preventing circumvention, they seem to be opening up that ability to evade ECOA, and I don’t understand why they would do that.

Chris Willis (20:47):
I agree with that. And I think, as I said before, that is not just true for advertising, but true for a variety of fair lending issues where we could see a return to the very high degree of focus on fair lending that we saw during the Biden administration and then even have it expand into other areas. Redlining was a huge focus area for the administration during the Biden administration, and we might get that back. That’s what we talked about last time you were here on this podcast. But there’s nothing to limit that just to mortgage redlining. It could be five other issues as well. And I think industry correctly perceives that that change in attitude is likely to happen and that our compliance decisions have to be based not on what the regulator’s view is today, but what it may be tomorrow, because our business model calls for us to stay in business for more than the next two and a half years.

Stephen Hayes (21:33):
I cannot say with any confidence that it’s a prevailing position among irresponsible lenders, and there are many of them in the market. And so the idea of giving them this permission, I don’t understand why the CFPB would do that. The rationales they gave, I think, are very difficult to follow. But I am worried about this compliance slippage in this moment. It’s pervasive.

Chris Willis (21:54):
Yeah. Well, you mentioned the litigation that’s pending in the District Court for the District of Columbia in which you are representing a series of plaintiffs challenging the CFPB’s rulemaking, the one that was finalized in April under Reg B. Since I have you here, I would really love for you to give the audience an update on what’s going on with the litigation and what should we be watching for.

Stephen Hayes (22:14):
Well, the short answer is that right now we’re in the middle of summary judgment briefing. The plaintiffs just submitted their motion for summary judgment and brief supporting that on the 29th, and that’s publicly available if anyone wants to see it. But there’s a set briefing schedule. So the government, the CFPB, is being represented by DOJ. DOJ will submit its own briefs, and we’ll do briefing back and forth through about mid-January, and then it’ll be up to the court to weigh those briefs, probably hold a hearing, and we’ll see.

Chris Willis (22:45):
And just to help the audience understand the briefing and the ultimate decision that will be rendered in that case, the claims that the plaintiffs have asserted there obviously include arguments that the rule as revised is not consistent with the statute. I’m sure those arguments are there, but there are other sort of more procedural arguments under the Administrative Procedure Act too, right, Stephen?

Stephen Hayes (23:08):
Yes, there is a whole slew of arguments. So there are arguments for each of the three sections of the rule arguing that they are contrary to the statute, that they are just unlawful. There are arguments that are what are called APA arguments, Administrative Procedure Act arguments, about failure to comply with the requirements of the APA, and then similar arguments related to failing to comply with other procedural issues, such as doing a cost-benefit analysis required by Dodd-Frank and a few others. So there are a number of arguments, both procedural and substantive.

Chris Willis (23:40):
Which means that when we get a decision in the case, it may hit the substantive issues of was this discouragement provision consistent with the statute? Does the statute allow for disparate impact? Can you restrict special purpose credit programs in the manner that the CFPB has done? But equally, we could get a decision that doesn’t touch those issues at all and just says, oh, there was some failing in some aspect of the manner in which the rule was promulgated.

Stephen Hayes (24:04):
Yeah, absolutely.

Chris Willis (24:05):
Yeah. So that’s, I think, an important thing for the audience to watch. And of course, I’ll be watching too very eagerly as the briefing plays out and as we then get a decision from the district judge sometime next year. Stephen, I really appreciate you being on the show to talk about these issues, and I think I want us to together leave the audience with the message that we were talking about a minute ago, which is we have to make decisions about targeted advertising that will be defensible not just in the view of today’s CFPB, but tomorrow’s CFPB as well. And there’s every reason to believe that the CFPB will become interested in this issue again. And you made a very good point, which is if they disagree with the discouragement changes that were made to Reg B, and surely they will, they don’t have to just put back the pretty general language that was in there from the ’70s. If they’re going and tinkering with the rule, they can make it much more prescriptive and much more aggressive, as you noted. And I think there’s a very good chance of that happening. And that, I think, informs the decisions that responsible lenders will make today about the manner in which they do targeted advertising. So I really appreciate you helping me make that point.

Stephen Hayes (25:06):
Yeah, absolutely. And if there’s one other point that I could make before we drop it, I would also, for any lender thinking about its compliance related to targeted marketing, I would think about it as a fair lending issue, as we’ve talked about, but I would also bring a pretty significant UDAAP lens to it, and not just in the classic UDAAP lens of what does the copy look like. We have started to see almost bipartisan concern about what some call surveillance pricing, but really it’s pricing based off of data that might be surprising to a consumer in some way. So we’re not really talking about risk-based pricing in credit, but we’re talking about pricing where someone might use a model to predict willingness to pay or something like that. And some of that goes on, but I’ve started to see in a marketing context it blend a bit. So I have seen marketing that does affect access to different pricing tiers, pricing levels, and it’s framed as a marketing issue, but it is affecting pricing. We now see even this FTC issued a proposed policy statement about concern that people getting individualized pricing in ways that are surprising could be a deception risk, a UDAAP deception risk. There was a hearing in the Senate where bipartisan senators agreed that personalized pricing in ways that might be surprising or unfair would be a significant issue. And a lot of states have focused on this too. And so to the extent that advertising marketing models start to bleed into underwriting and particularly pricing, I would really keep a UDAAP lens on that as well.

Chris Willis (26:41):
I think that’s a great observation and a very important thing for the industry to keep in mind. So Stephen, thank you for raising that and all your other comments on the podcast today. I’m so grateful to you for being on the show today, and of course, thank you to our audience for listening as well. As I said at the top of the show, please visit our new combined financial services hub at troutmanfinancialservices.com where you’ll have access to all of our content, all the blogs, all the podcasts, and everything. And while you’re at it, why not visit us on the web at troutman.com and add yourself to our Consumer Financial Services email list. That’s the list we use to send out our periodic alerts and advisories, as well as invitations to our occasional industry-only webinars. And of course, watch your podcast feed every Thursday afternoon for a great new episode of this show. Thank you all for listening.

Copyright, Troutman Pepper Locke LLP. These recorded materials are designed for educational purposes only. This podcast is not legal advice and does not create an attorney-client relationship. The views and opinions expressed in this podcast are solely those of the individual participants. Troutman does not make any representations or warranties, express or implied, regarding the contents of this podcast. Information on previous case results does not guarantee a similar future result. Users of this podcast may save and use the podcast only for personal or other non-commercial, educational purposes. No other use, including, without limitation, reproduction, retransmission or editing of this podcast may be made without the prior written permission of Troutman Pepper Locke. If you have any questions, please contact us at troutman.com.

DISCLAIMER: This transcript was generated using artificial intelligence technology and may contain inaccuracies or errors. The transcript is provided “as is,” with no warranty as to the accuracy or reliability. Please listen to the podcast for complete and accurate content. You may contact us to ask questions or to provide feedback if you believe that something is inaccurately transcribed.