Get the Monday email →
Loading player…

Rewriting the Rulebook: Rethinking Regulation and Engagement | DC Fintech Week 2026

DC Fintech WeekTue, Nov 3, 2026, 12:00 AM

Financial rules are traditionally developed by governments and applied to private actors through siloed, often fragmented processes. But rapidly evolving digital markets may require a more collaborative model. This session introduces the Fintech Regulatory Futures Index and explores how regulators, firms, and other stakeholders can engage earlier and more effectively — and how digital tools can reshape consultation, notice-and-comment, and the rulemaking process itself. • Moderator: Chris Brummer, Chief Executive Officer, The Fintech Foundation • Todd Fox, President, Visa Economic Empowerment Institute • Peter Kerstens, Advisor, European Commission • Jérôme Beaumont, Executive Secretary, Egmont Group

Transcript

I am absolutely delighted to have an all-star panel. On the one hand, we have Jerome Bumont who is leading as executive secretary of the Egmont Group. When you think about an organization like FATF, the financial action activity task force, thank you very much. And for any money laundering and AML KYC issues, there are financial intelligence units who support these operations globally and this gentleman here is the point person. We have next to him Peter Kerstens who is not to mention one of my very good friend, so I want to make fun of him, but in a serious capacity an advisor to the European Commission and is one of the most senior officials in the European Commission when it comes to really all areas of digital asset policy and innovation.

We have Yesha Yardley of making she knows everything due to me over over at Vanderbilt and a very close colleague to offer her insights again. And then we have Todd Fox. Todd is the president of Visa Economic Empowerment Institute and both Yesha and Todd sit on the board of the Fintech Foundation which is bringing to you DC Fintech Week today. So thank you to you both. Okay, first of all, you know, there was a very interesting again, I had not heard any of these comments, but but the initial comment about the agentic almost sort of agentic waged sort of attacks against regulatory agencies and other sort of public infrastructure.

Found that to be kind of interesting. I had I had not I am not familiar with any of that. Mr. Kerstens, I mean has this crossed your radar? Like what is the scope of concern for sort of let's call it agentic malpractice and also sort of state directed agentic I don't want to say warfare, but but agentic harms that can be inflicted against regulatory agencies, has this ever popped up on on your radar?

It hasn't popped up for us, as far as I know, towards us, but I don't know all the attacks that happen on us in these times of super intelligence, because that's what they're what it's all about now. Um, but we are seeing super intelligence based motivated attacks on financial infrastructure. And that is causing a lot of concern and harm, because it puts artificial intelligence, super intelligence, whatever you want to call it, into the bad light. But, what we already see is that in order to defend against that, you'll also need artificial intelligence, super intelligence attacks. So, yes, this agentic behavior supported by artificial intelligence, super intelligence, is really changing the ballgame for regulators, but also for financial institutions, and also for the regulation itself, because if you look at legislation and regulation, it usually is addressed towards people and entities, legal persons.

They have to comply with the law. What does it mean

For the love

if an agent starts operating? People say, "Well, someone will control the agent." From what I can see is that, "Well, sometimes some No one seems to be controlling anything." So, what do you do then?

So, um, I'm going to sort of jump, although I see I see eyebrows from Todd Fox, and sometimes his his eye I I can't I can't read the eyebrows correctly, but certainly, you know, Visa's at the forefront of lots of agentic commerce. But, you know, in terms of this kind of you know, we'll be over at FINRA, we'll be talking about agentic fraud, frankly. But, but are are any of the concerns there, including this idea of showing the receipts? I thought it was an interesting comment, sort of in the next Congress. You know, do do you see any kinds of threats in your world that sounds familiar?

And and and and what what what kind?

Well, Chris, thank you for the question. And yeah, I thought that what Chairwater said in the panel immediately before is is very important. I think everyone that interacts with agentic AI goes through a version of the same stages. First, it's it's this amazing magical technology that can go out and do things for me, but it's it's it's largely kind of packaging information, giving you answers instead of directing you to a place where you have to find it yourself. Then you start to realize, wait a second, if the bad guys have this technology, then that could be weaponized against my data.

And suddenly, there's a there's a there's a there's a moment where you start to realize, oh no, I'm going to have to actually shrink the attack surface and harden my defenses. Last year at DC Fintech Week, I sat on the stage and talked a little bit about how Visa was building something called the trusted agent protocol because there were so many merchants in the world that needed to figure out how to tell the difference between an agent that is actually their new customer and an agent that is sent there to to wreak havoc on their on their e-commerce environment and and breach their data. So, where are we now? I think the fact is we have seen some progress in the adoption of agentic AI for commerce and for payments, but not as much as I might have predicted a year ago. And the reason why trust is still the big unsolved problem.

If you look at surveys that we've conducted, we ask people, you know, do you trust an agent to execute a payment on your behalf? Around 20% of people say yes, which means the vast majority of people still really don't trust what we would describe as almost fully autonomous agentic payments. But if you ask if it's a Visa transaction, well, then the numbers go up pretty considerably. It's it's over 60% for the general population and it's over 70% for people that are active users of generative AI. So, trust is the big unsolved problem.

It's an arms race. You need to have the make sure that the good people with AI are innovating fast enough to stay ahead of the of the malicious actors and this will continue to be the big defining story in cybersecurity and payment security and fraud for years to come.

Can I just add to that?

Please.

Um so as Todd mentioned, like this is hard for the marketplace to get its arms around. It's also super hard, as you mentioned, Chris, for the law. So, there are a couple of really interesting problems as Peter alluded to as well that really stand out here. Right? Firstly, there's no real legal category for these agents, right?

Like they don't have a specific legal personality that we can build a legal system around and say, "Hey, you're responsible for that." So, even if we want to build sort of agentic protocols and standards, we have to figure out a new category of legal personality that can attach to it. The other part of this is that often times tracking, even though we have data, can be super hard. Some of these wallets, they are phantom wallets, meaning that they come and they disappear. Right?

So, they make a transaction or they do a deed and they just kind of disappear. How do you track that? How do you follow that? It makes it harder for regulators to deal with. And the third part is that liability loss allocation.

When these things don't have a legal personality, when they're kind of acting and doing interesting things, when they can potentially disappear, just come and go, then who's going to bear the risk of the loss? Is it going to be the folks we can identify like the Visas and the banks and others? We've worked super hard to create real rules of the road for liability, but now they're going to be really tested because no one wants to take on the open-ended potential liability that might arise in this case. So, the law is going to struggle here in the short term.

So, so I would I would I would ask like Jerome about these questions, but I'm always afraid that like you know, it's like you know, ask the wrong questions, you know, I tell you and I have to kill you only kind of sort of. So, you know, instead of getting into the the specifics, although if you do have a sort of agentic sort of ideas and in terms of, you know, what that means for the FIUs, I want to sort of move a little bit with where Yesha was was taking us when it comes to the international standard setting. Over at the Fintech Foundation with a lot of work with so many people including Yesha as as our co-chair and with and and and with Todd, um you know, we we launched sort of our own initiative uh really working around the world and with and with Peter Curzens as well. Peter, you you you also contributed to that to that to that effort. Um uh you know, and when you think about creating categories, say, for agents.

When you think about standard setting at at a general manner, how how do those kinds of things um you know, when you have a a kind of a public private sector initiative with a with a kind of a global scope, but kind of operating um in a way that's that's quite literally begging for criticism um publicly, I it kind of differentiates itself from how, you know, your your the the official sector works with its international standard setting. What Brett McIntosh bring up the FSB and the others. Do these pro- How does the official sector the process of making standards in the official sector for things like agentic threats, is that up to snuff? Like how does it operate when some of it is private, some of it is public? Should more be private in order to make things move faster?

Should there be more transparency because you need more brains and minds involved? Just What kind of diagnosis, kind of vision do you see after your years at the Egmont? Um, you're sort of looking at standard-setting international.

All right. It's a long question.

I could have just said, "What do you think about like these folks attacking the states by the way?"

That that would have been not appropriate. I will try to be concise in my in my answer. Um, having been attending FATF plenaries for like 15 years, what I noticed is some years ago basically the delegations, the members of of FATF would be designing the international standards on their own and and with a very little say for the private sector. So, [snorts] heads of delegations and their delegations mainly treasury, uh, sometimes central banks, uh, and and ministries of finance would would basically design those international standards. Uh, more recently, uh, interestingly, FATF has been opening up a lot to public consultations.

So, give you an example, uh, this year and the deadline is was August. So, all through the so from spring to summer uh, there was a public consultation on recommendation 16 which is a payment transparency {slash} travel rule where there was a public consultation and and the private sector was heavily invited to contribute. Last year it was on financial inclusion, uh, the previous years on the to combat the abuse of NPOs for example. So, I see more and more an opening when it comes to standard setting an opening to to the to the private sector or to the public more broadly. So, I think it's a it's a good development.

It brings some challenges obviously. First, the more you are around a table to draft something, messier it is. It takes more time. This is one. Second, there is a an issue of confidentiality to a certain extent, since some bad actors actually would love to be involved in in the design of those recommendations to for their own benefit.

So, it comes with there are opportunities and challenges, but I really see a shift or at least a curve in in the way of international standards, especially FATF standards, are are being designed.

So, Peter, you are at an international or regional organization that operates in an international system. When you look at, again, and I'll ask a sort of an iteration of the question that I asked Brett, I mean, you think think quantum, think agentic, you know, we we have privacy issues that have of long been of of of of of, you know, the importance in the European Union, but international standard setting, is it going to be fragile or fragile enough? Oh, I was slip. Is it going to be engaging enough, flexible enough, innovative enough to move at, if not machine speed, at least a practical speed, when you look at all these new opportunities, but also threats?

Well, I think the short answer is no. It's completely ill-adapted to the developments that we currently see, both national standard setting, regional standard setting, and international standard setting. And of course, we can consult industry, and we consult industry all the time, the private sector. And our most recent experience is that we basically consulted AI because of our mega consultation, which we just concluded, most of the replies I read were generated by AI, the large language models. Um, we then use AI to analyze it, sort of the AI is analyzing itself.

You don't get any wiser. What you do get is you get more commonality slop, AI slop text. Um, but when consultation works and private sector input is solid, you get also more opinions and strong held views and I I'm acutely aware that I'm surrounded here, well, actually the majority on this panel are law professors because you're a law professor, you're a law professor, I'm a fake law professor, so that's three. And law professors seem to want to categorize everything. Let's have a definition.

The problem needs a definition and then once you've defined the problem you have a solution. Well, the lawyers think they have a solution, but no, that's just the beginning of solving a problem. With agentic behavior is that or with quantum, whatever, it's a very much a moving target and I think it is very dangerous to try to define now what we don't know how it will develop because the definitions which will come up now will describe the realities as they are now. And with a bit of effort, we might be able to predict what's going to happen next year, in 2 years time, but further ahead I wouldn't venture that. But if we set definitions now, we're going to be stuck with them for a long time.

Just like

[snorts]

old definitions of what is a legal entity, what's a legal person and that worked for a while until it no longer worked. So, I think it's extremely difficult to set standards and therefore maybe we shouldn't.

Well, you know, that was part of the tongue and cheek of the project in general with with with with the RegEx, which is to really instead of a perpetual future, you have a perpetual standard, right? So, at any point in time, you can basically, you know, someone has a better idea, okay, great, you know, people are voting it up in, you know, kind of a crowd source version a kind of um you know, a certain adding a certain velocity of one will to the international standard setting process, which is which is which is kind of um new and and interesting. Uh just even getting to the project, but I will say as soon as you mentioned to categorize the agent, I looked at Peter. I was like, Peter's not going to agree with me.

[laughter]

And like, you know, as as as soon as you said that, you know, it's it's you know, we have principles basically Um and so, feel free to to to respond, but you know, just just looking at that challenge and even the experience of sitting around in the working group with all these people, some of them in Australia up at 1:00 in the morning, some like Port Todd Fox up at 6:00 in the morning, 5:00 5:00 5:00 in the morning once on these on these groups. There was a lot of effort put into the Canon. Um you know, like like how do how do you move uh international standard setting? I'm asking you Yasha, because so so that you Yeah, so you can respond to you know, Peter disagreeing with it.

So, I I have a So, this is why I love the RegDex project because it forces accountability on the international community to do something. We've really lacked standard setting around FinTech and around all of these great dynamics that are coming up in finance like AI, proper you know, quantum, etc. And now finally we have like some global experts that are saying, "Hey look, we've come up with a methodology. You can't ignore it. Now let's react to it."

Now one thing that I was super worried about before the RegDex kind of came up was that even though we might have produced international standards on like banking and derivatives in the past, we lost incentive to do it on tech because all of us are competing on tech. We heard that in the quantum piece, right? China and EU and US we're all kind of kind of fighting for what feels like a zero-sum game. But, at this point we also know as Todd said, we need to do something. We can't look away.

This is happening whether we like it or not. And so, that is why RegDex is so powerful. Here's something, if you don't like it, have at it. And so, let's kind of move on and come up with something. And that's why I'm like super passionate about this thing.

I would just add to that and say there is far more interest in good faith dialogue between and among the different participants in building the future of finance than is typically reported in the press. I think the consensus narrative is that we're in this very fractured and fragmented world where no one trusts anyone else and it's all geopolitics. The fact is the very existence of this panel shows that we actually are in a moment when there is a genuine desire to exchange views between the smart people from the official sector, the builders in the private sector, and the academic community. And what you have built with this uh the Fintech Regulatory Futures Index and with DC Fintech Week overall is the venue where we can all come in in good faith and and check our bias at the door and say, "Look, you know, we can't make this work unless we're all working together from first principles." And so, while I agree with Peter to get 200 countries and territories to agree on like a new standard uh we do have to have someplace where we can talk about interoperability.

Because the tendency of all these new technologies is toward global interconnectivity. If the technology is connecting us, but the policy environment is not interoperable, then we're just going to enlarge that attack surface and set ourselves up for more of the problems that we're trying to avoid. Uh the bad guys love a fractured and siloed world, so we need to have some kind of interrupt to to make it all work together.

All right. Well, it was a Peter

No, no, no.

I If I can say one quick thing, which is thank you, Chris, for an amazing day and for bringing us all together for a fabulous conversation. So, thank you for that and surviving

up until It's not done. We have Well, thank you very much. Uh and thank you to the panel.

Automatic transcript — names and jargon may be misspelled.