The Architects: Reimagining The Financial Future

The Finternet: GenAI, Tokenization, and the Future of Public Infrastructure

• TWIF

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 56:36

The views expressed today are not necessarily that of the Federal Reserve System or the New York Fed. 

This week on The Architects with Flourish hosts Julie VerHage-Greenberg and Kabir sit down with Siddharth Shetty, CEO of Finternet Labs, and Michael Lee, Financial Research Economist at the Federal Reserve Bank of New York, to explore what happens when payments are no longer limited by human effort.

Instead of users choosing banks, rails, currencies, or routes, AI agents can search across liquidity pools, cost structures, and time constraints in real time and execute the optimal transaction on your behalf. Faster, cheaper, and designed around outcomes rather than effort.

From the Finternet to tokenization and regulatory design, the conversation examines how programmable money and intelligent agents could reshape financial infrastructure and expand real financial agency at a global scale.

If finance is becoming programmable, this episode looks at what that unlocks next.




Flourish Ventures is an $850M global early-stage venture firm that backs entrepreneurs transforming financial systems for the better. Its portfolio spans more than 100 companies across the U.S. and emerging markets. The firm also supports innovators shaping policy, media, and research to accelerate lasting change in financial services.

This Week in Fintech (TWIF) is the largest fintech community in the world, presenting news, podcasts and newsletters from around the world.  Subscribe to our newsletter to stay up to date on the latest in news opinions, and all things financial technology.


Agency As The North Star

Siddharth Shetty

What we've seen is if you design technology with the right kinds of regulatory frameworks, you can really use that to dramatically give people agency. And that's very critical in the world that we're heading to. People should have agency over their own digital identities. People should have agency over broader credentials that surround them that they ask, where did I study? Where do I work? How much do I earn? People should have agency over their own flow of money. I should be able to transact with whom I want when I want. You know, I want to transfer money to an individual, I want to transfer money to a business. A government wants to transfer money to me in the times of a crisis. They should not have to send me physical checks to my email, right? And this could apply to both individuals as well as people, small businesses and enterprises. And so fundamentally, it's about giving people aid.

Series Context And Big Idea

Julie Verhage-Greenberg

Welcome back, you guys, to the This Week in FinTech and Flourish podcast in this special series that we co-created with Flourish Ventures. We're exploring how emerging technologies from AI to tokenization are reshaping the financial system itself.

Kabir Kumar

Today we are talking about something that could redefine financial infrastructure for the next generation. It's very exciting. We're going to use the word Siddharth uses the Finternet. Imagine an internet for finance, one that's open, programmable, and interoperable by design.

Julie Verhage-Greenberg

And I don't think I had heard that term before. So it's something new to me as well. I love it. So joining us are two guests at the heart of this transformation. Siddharth Chetty is the CEO of Finternet Labs, hence the word Finternet, and one of the architects behind India's open financial ecosystem and from the account aggregator framework to what he now calls the Finternet, a public good for global finance.

Kabir Kumar

Great to have you, Siddharth. And Michael Lee is a financial research economist at the Federal Reserve Bank of New York, where he studies digital money, payment systems, and how innovation intersects with financial stability. He's written some really compelling research papers over the last two years that are effectively, I would say, the intellectual underpinnings of the Finternet.

Meet Siddharth And Michael

Julie Verhage-Greenberg

And together we're going to unpack what a truly open, tokenized financial system might look like and how to balance innovation, privacy, and trust as we build that. So let's dive in. Amazingly, so uh Siddharth, since you turned have the term FinTernet, let's start with you. Uh, you've been at the forefront of the open financial infrastructure from India's account aggregator, like we mentioned, um, to now the FinTernet. What problem are you solving with the FinTernet? And how does it build on what you learned from India's open financial experience?

Why Finternet As A Public Good

Siddharth Shetty

Thank you so much again for inviting me. And it's exciting to do this with all of you and be a co-panelist with Michael. Um maybe setting a bit of backstory to how we arrived at the Finternet. We've spent the last 15 years designing and building digital public infrastructure. An early part of that journey was in India, but in fact, the last five years, we spent time working with about 30, 40 countries across the world. And, you know, through that journey, we built a bunch of things. You know, the world's largest digital identity system, the world's largest retail payment system, uh, then the open finance ecosystem as well. But what we realized was that the way we build financial infrastructure is very purpose-specific. So it's specific to a country, specific to a sector, specific to a use case. And you see that across the world, right? If you look at fast payment systems in different jurisdictions, um they have similar attributes, but when you start to try to link them together, these are often five-year efforts with complex bilaterals, which you know, imagine enabling that for 190 countries plus, you're talking about beyond our collective lifetimes. And so what we realized was that traditional financial infra being so purpose-specific, there had to be a better way to architect these systems. And um, if you think about it and you go back to you know, pre-digitization when all of us had physical tokens. Let's say I met you, Julie, and you know, handed you the US dollar bill, um, I gave the US yours. And that was a self-describing, self-contained token. And when we digitized it, what happened is now my bank had to integrate with your bank and all the complexities came in there. Uh, but if you look at the newer cryptographic developments that are coming in, public blockchains, tokenization, etc., it's essentially a version of Back to the Future. So you can get that kind of portability, that kind of interoperability uh through representing all of these asset classes as digital tokens. And so that's really what drove us to then put together the Finternet, uh, which was a paper that we co-authored along with Nandankenny, Pramod Bharma, and Augustine Castins and released this in Washington, DC last year. Um, and then we've spent time essentially designing and building it out, which we'll get to, I'm sure, in the in the latter part of today's podcast.

Kabir Kumar

Siddharth, um, you know, I had the good fortune of being involved in the early days of what eventually became the India Stack. And we were just uh a bunch of us, uh, as Julie was saying, we were in at Money 2020 last week and we were reminiscing the kind of the launch of UPI and it w it was in a ballroom in a hotel, there was a disco ball, and everyone was very excited because it was this amazing piece of uh infrastructure that was being uh released effectively. If internet seems to take it a step further, it takes it beyond national boundaries, as you were uh intimating. Uh help us understand why it is a public good, you know, what does it actually mean in practice? And then how do you design it in in that in that sort of uh with that outlook? How do you design it so it serves both the public interest and and you know serves commercial innovation?

Siddharth Shetty

So essentially that partnership is very critical. Uh public infrastructure is the base, but private innovation that gets unlocked on top, because if you don't have private innovation, you'll never really get to scale. Um and you're not going to bring in the innovators who drive and make that scale happen. Um and similarly, if you take a look at UPI, which fundamentally was a protocol that allowed you to transfer value from one store to any other, uh, but it got implemented on traditional Rails. So it was only settling INR fundamentally. Um if you design that protocol today, you'll end up again looking at something similar to the Finternet, which works across multiple assets. But um, going back to UPI, what it did was you had an open protocol that allowed you to transfer value in this interoperable manner, any bank account, credit card, wallet on one end, and any consumer app or merchant app on the other end. And across this multi-party system, you could have a payment transaction done in a safe two-factor authenticated manner. And so if private innovators came in, built on top of the protocol, raised billions of dollars, acquired, you know, tens of millions of merchants on one end and hundreds of millions of consumers on the other. And so that sort of partnership is very critical. Um, and so similarly, when we look at the Finternet, uh, there are critical infrastructure blocks in the internet today that have to be accelerated. So, how do you establish trust in a world of tokenization and agentic flows? Um, other key building blocks that we're building out within the labs. But we ourselves, as well as other entrepreneurs, would go on to set up operating companies uh to further drive this and bring this to life, um, uh, especially at a global scale.

Economics Of Platforms And Tokenization

Julie Verhage-Greenberg

Michael, you are the economist in the group, and I love that you have like this big whiteboard behind you. We can't see what it says, but I feel like there's like economic knowledge just sitting back there, ready to like turn this world around in this tokenized future. Um I got very familiar with the Fed's payment systems, both from reporting, but once after I was a reporter, when I moved more into the marketing side, I worked at Orm and they were one of the first companies really working to um enable Fed now. Um so that's something that I got very familiar with. Obviously, like that was before we were really talking a lot about tokenization and stable coins and all these things. How does this all work together? And what do you think sort of like that roadmap looks like from your seat?

Michael Lee

It's a great pleasure being here. The views expressed today are mine and not necessarily that of the Federal Reserve System or the New York Fed. I think there's a tendency in common discourse to take a rather reductionist view on the features and characteristics of today's financial architecture. And I think this kind of ignores the fact that the current financial system, it's it's not an artifact of a single mastermind or even a small group of designers. Its shape and form have been determined over time through a confluence of forces, shared experiences that helped us build on top of it. And so, in this sense, I think it's the more appropriate way to view it right now is that this is part of the evolutionary progress of what we think we can achieve through a better design financial system. And if we look back, uh, this means that even though some of the technology and innovations that we're seeing are revolutionary in nature, if we take a step back, this is really part of the progress that we hope to achieve as we're trying to improve the efficiency and stability of our financial system.

Julie Verhage-Greenberg

That makes sense.

Michael Lee

That makes sense.

Julie Verhage-Greenberg

How familiar, like I said, I hadn't heard the term finter net before. Had you heard that term, and how familiar are you with that term? And like what are some of the takeaways from like what Siddhartha was saying with the India's financial ecosystem and how money moves to what would and would not work here in the US for various reasons?

Futures For Our Kids’ Money

Michael Lee

Great. So I was not familiar with Finternet before, but I think the word itself really captures the essence of what we aspire to build, not just in any particular jurisdiction, but like globally. I think there are several kind of key aspects to think about when we're thinking about uh infrastructure, though. First of all, I think it's really important for us to distinguish between the types of monies that we're trying to transact and the actual transfer technology. Uh I have prior work with uh colleagues really thinking carefully about what we're trying to actually accomplish. And then we need to be really careful about the technological features that enable us to achieve new forms of arrangements or partnerships versus the actual assets that we care about, whether it's pertaining to public money or if it's uh a claim or a security. But one of the things that I think is more important as a as a principle when we're thinking about public infrastructure is really the commitment as a public infrastructure. We know that there are huge economies of scale, and these are all inherently platforms, and platforms always have this huge temptation to extract rents once they are established. And I think this is a key limitation and challenge that any private or public sector uh development uh faces when they're thinking about how do we create an equitable system that allows for the ownership and the rights and the economic gains to be distributed in a fair way. The third thing I'll say, and we have the benefit today of over a decade of huge experimentation uh in the blockchain space. And I think what one key learning there is that there are significant complexities to having emergent economic behavior to actually match the design intentions. And so whether it's a blockchain or a protocol, the work is not just in terms of the technological choices, but actually the economic mechanisms that are wired into the fabric of their design. And to make sure that there is a synchrony between the way in which we intend behavior to work and how the actual systems work. I think this is something that uh more and more technologists are appreciating and integrating into the thought in terms of how they design these systems. Uh, we're thinking about a much more complex and sophisticated arrangement uh that hopefully, from the user perspective, is simple. And that requires a lot of dedication in terms of thinking about the economic design, not just in terms of the technical design.

Julie Verhage-Greenberg

Okay, yeah, yeah, yeah. That makes sense. And you mentioned blockchain too. It's been around for a long time. And for the longest time, I remember hearing people like, oh, yeah, it's going to be the next big thing. And it's really taken a long time. I feel like, you know, figure going public and kind of helping people understand it a little bit more is one step in that direction. But you and I were talking before we started, and we each have daughters that are around three years old. What do you think, you know, the financial system will look like by the time that they are old enough to be transacting this money? So I want Michael to answer this. And then I would love Siddhar for you to answer from um the India point of view too. I don't know if you have any kids or anything, but if you did have a young kid, what would the financial system look like for them in the future?

Michael Lee

So I actually think that future, we have a lot of agency. We need to make the right choices collectively in order to ensure that it has the values and attributes that we think they deserve and the future generation deserves. And fortunately, we're at a standpoint where we have the underlying technology, if used and combined correctly, can actually achieve what a desired financial system is. It's a system that's equitable, it's a system that grants rights and ownership properly, but also is efficient and safe. And I think these are all general attributes that we think uh should be uh accessible to everyone. Now that's it's easy to list a set of desirable attributes, and I think part of it is to not lose sight of the big venture that uh everyone in the space is trying to should be trying to achieve, which is to make a financial system that's safer, more efficient, that's seamless, um, in a way that we can depend on it for many decades, if not centuries, to uh uh afterwards.

Verifiability And Transactability Defined

Siddharth Shetty

Siddhar? So, in my view, um I think there are two parts to it. Um and I do want to clarify that what we're doing with the Finternet is not just about taking what we've done in India um and applying that outside because that fundamentally doesn't work. Uh, we do draw on a lot of our learnings because obviously, how do you design infrastructure at population scale? Kabir has been part of that journey. How do you design it in highly diverse, heterogeneous environments? How do you traverse multiple devices? How do you do interoperability? So there are interesting learnings there, but there are a lot of other things that we're now applying in a more global context that are different than what we did purely in an India context. But but if I answer your question, I think what we've seen is if you design technology with the right kinds of regulatory frameworks, you can really use that to dramatically give people agency. And that's very critical in the world that we're heading to. People should have agency over their own digital identities, people should have agency over broader uh credentials that surround them, that they cast, where did I study, where do I work, how much do I earn? People should have agency over their own flow of money. I should be able to transact with whom I want when I want. You know, I want to transfer money to an individual, I'm gonna transfer money to a business, a government wants to transfer money to me in the times of a crisis, they should not have to send me physical checks to my email, right? And this could apply to both individuals as well as people, small businesses and enterprises. And so fundamentally, it's about giving people agency and agency over beyond their money flow as well. Um, as people start to then transact, I should have agency over my data so that I have a unified view. I can start, you know, figuring out how do I improve my own economic life. I should have agency over the ability to raise capital. Today you have very high transaction costs and equity markets that exist. The new kinds of infrastructure will completely democratize the concept of ownership to individuals around the world. So I should have the agency to participate in the upside of new technologies, new markets that are being created. It should not just be, again, restricted to the top 10%. And so I think fundamentally, if you look at the new infrastructure that's coming out, um, what we're really excited by is if you put the sort of Lego blocks together well, you can massively democratize access to billions of people around the world. And that then gives them agency, that gives them a sense of empowerment, that allows them to dictate, you know, how they want to live their life, what are the values they have, um, who they want to transact with, how they want to grow, how do they want to protect themselves uh through different kinds of financial instruments, um, and so on and so forth.

Kabir Kumar

Yeah, I think both of you, uh Sudhat and Michael, are saying that technologically a lot more is possible today than it was before before in terms of if you want to design a more fair system, a more equitable system, a more inclusive system, but it comes down to the design choices we make and the politics and the governance decisions that we make. I'm interested, Michael, to your comment at the beginning when you said, look, you know, we we tend to be reductive. Uh this has been a long-term evolution, lots of trade-offs and choices. I'm curious uh why uh this sort of area of tokenization, programmable money, why is it an important area of analysis at the Fed? And you know, why why are you focused on this? Why is this important for the Fed? Uh help us understand how it connects with sort of the existing financial system, um, you know, and and your kind of how you see the evolution happening within the Fed environment.

Julie Verhage-Greenberg

I want I want to add on to that too, with your answer. Is there do you feel like there is a bit of a consensus at the Fed as to what it will look like? Or do you think that there's a lot of different varying viewpoints points as to what the the future could look like with all of these things?

Stablecoins, Access, And Legal Guardrails

Michael Lee

This is a this is a great question. And I think uh this is uh a common question that I get when I speak to people outside of the regulatory sector. And first, we tend to think that as a regulatory institution, the Fed or any other policy institution for that matter has a very singular view in terms of how the future looks like. And in fact, uh what we actually value is a very diverse and heterogeneous set of views. People have differing preferences, differing philosophies, and even uh ideas on what the future looks like. And that's how, even internally, uh, we are able to uh have a more well-rounded view of what we should be doing, what we should be caring about. Now, having said that, with the advent of distributed technologies and uh blockchain, there is massive amount of effort across the Federal Reserve System in terms of experimentation and studying the technical and economic properties of this. And so uh, you know, part of it is that in as a public uh institution, we deeply care about progressing the future of the financial system and making sure that there is the right type of support and uh safeguards required for it to succeed. You know, naturally speaking, that means that there is significant public private partnerships. But more importantly, I think in order for for For the engagements to be effective, it also requires a lot of internal human capital in terms of understanding the technologies, the limitations, and the strengths. Tokenization is one next step that we that I personally view as one way to broaden the scope for different types of financial arrangements and also to allow for multiple different parties to achieve arrangements that was really challenging to do in the current uh legacy environment. Having said that, in the end, a lot of this is also an opportunity for us to reconsider how the regulatory and legal frameworks can be adjusted so that we can achieve the efficiency gains and also achieve some of the regulatory and legal principles that we want to achieve in a financial system, given that now we have a different set of tools. And I think this is the conversation that we should have. Not just criticizing how the system works or how it's implemented today, but being thoughtful about retaining and even enhancing some of the principles that we want to guard. There are reasons why we have certain types of regulations. And there's a reason why some access to actions or services are restrictive. How do we achieve a better world through the technology instead of trying to append the whole system and replicate it from scratch again?

Kabir Kumar

I really appreciate that. And you know, because at the end of the day, partly why we've struggled with sort of where blockchain goes is because we've been sort of trying to find a use case for the technology, it seems, collectively. You know, this is exciting, this is interesting. What what what what problem can we solve with this? And I think now we might be in an era where we are starting with perhaps more false principles, uh, we are we're starting with a vision for where we want to take the financial system, the technology seems more mature, the community around the technology seems more mature. So you know, the ambition that's expressed in the FinTech feels more graspable to me than it was before. And that's what's exciting, I think, and that's what both what both of you are saying.

Ad

If you're building in fintech, you already know stable coins are transforming the way money moves globally. But integrating them? That's a mess of wallet infrastructure, security ops, and compliance overhead, most teams aren't built for. That's where CrossMint comes in. CrossMint is the all-in-one platform that lets you launch stablecoin payments, tokenized rewards, and digital wallets through simple enterprise grade APIs. Whether you're building a neobank, remittance product, or global payroll platform. CrossMint handles the heavy lifting from key management to gas fees to regulatory guardrails, so you don't have to. Global brands and financial institutions like MoneyGram, Papaya Global, and Sedexo already trust CrossMint to bring real-world use cases on-chain. From cross-border stablecoin wallets to tokenized employee benefits and loyalty programs, and with infrastructure that spans 40 plus blockchains and supports both human and AI agents, CrossMint is uniquely positioned to power the next generation of programmable agentic commerce. Start building with the infrastructure trusted by the best. Visit CrossMint.com to learn more.

Kabir Kumar

I think we want to understand the technology underpinnings a bit better. So to the extent we have time on this conversation. And I, Julie, you had a set of questions to get us started.

Julie Verhage-Greenberg

Um, Siddharth, I'd love to go to you. So I feel like I'm on a call with three people that if we were talking about payments in India, would be on like a payments 801 class, and I would be on a payments 101 class. So explain to me, like just understanding like what Michael was talking about and what we're doing in the US financial ecosystem and infrastructure. What are some similarities and differences to what we're seeing in India and how the Finternet is playing a big role in uh that as well?

Kabir Kumar

With the recognition that the Finternet is not just about India. We we, you know, that it comes out of the experience of India, but with that recognition, uh, we would love to understand the sort of technological uh underpinnings of how you're building this.

Privacy Tech Meets Compliance

Digitized Vs Tokenized Assets

Siddharth Shetty

Yeah. No, it it does go at times go back to India just given the scale of implementations there. Uh but we spent the last five years working with about 30, 40 countries, and a lot of that experience also informed, you know, you look at the developments of PICS in Brazil, uh, PayNow in Singapore, PromPay in Thailand. Um, and so a lot of those learnings have also informed how we've been thinking about the Finternet and how that shapes up. Uh, but fundamentally, um, there are a couple of aspects to this. I think one, if you look at broader tokenization, um, you're seeing it across different asset classes. And I think this will be a continuum. So we're at the start of what's fundamentally a 10, 15 year journey. So you have tokenization being applied, whether in the regulated space to different forms of money, central bank money, there are efforts like the BIS project agora. It could be private forms of money, whether it's tokenized deposits or stable coins. Um, it could be securities, you know, the SEC is now putting a sandbox together for that. You could have tokenization of registered assets. So these are assets that have a registrar backing it up, like your physical house, or um uh it could be, for example, a vehicle. You could have tokenization of attested assets. And so these are assets that have third-party ratings, attestations attached to them, for example, commodities, energy resources, and so on, or natively digital assets that are there. And so, really, tokenization can be applied across the entire universe of assets that we see. But what's foundational to it is really two key properties. One is the property of verifiability, which means that you know, let's take any of these assets. Uh today it's very expensive to verify is that indeed the owner, is that indeed the holder, who are all the owners behind it, if they're multiple stakeholders, what is the provenance of this asset? It's often physical paperwork, uh, or even if it's digital, it's really expensive to verify. Um, that's a common problem across the world. So the first thing that we're solving for is verifiability of these identities, credentials, and assets, so that any other counterparty across the world can know it is authentic, who they're transacting with, the asset that they're transacting with is authentic, the history is authentic, and so on. Uh, that's one aspect to it, but a very critical aspect because you can use that to satisfy your AML, CFT norms, you can use that for sanctions cleaning, you can use that for other compliance flows, like is this individual an accredited investor or not? All of those are downstream of being able to produce verifiable proofs about who you are and verifiable proofs about the asset. And the moment you can commoditize that, which is what technology allows us to do, digital signatures, wallet infrastructure, uh, and the like, uh, that's all moving in in that direction. So that's one part, verifiability. But that's not enough. You know, once I verify an asset or I verify ownership of the asset, or I verify the individuals, the next thing you want to unlock is transactability, the ability to transfer. Now, traditionally, transactability, as I mentioned in the start, was limited to very purpose-specific systems, right? So if I'm transferring US dollars to you, uh, essentially I needed access to the US dollar settlement system or via correspondent bank that had access to it, right? If I wanted to transfer security to you, I needed access to the security slash depositories infrastructure in a particular region, right? So what happened is each asset class that was, if I tried to transact property to you, that was again, in many cases, going back to a physical registrar of a local authority in different countries across the world to update ownership transfer, right? And I'm not talking about even more complex workflows like nomination, delegation, fractionalization, et cetera. We have some nuance around ownership because you don't technically own money in your account, it's a claim. And so uh we call it like holdership uh because ownership has a very specific legal connotation to it. But fundamentally, it's about being able to transfer property rights. Those property rights could point to ownership, it could point to other attributes as well.

Julie Verhage-Greenberg

Like transferring the claim to whatever that asset is, whether it be money or something else versus like transferring it.

Siddharth Shetty

Whatever that claims. So you have immutable infrastructure that allows you to make that transfer possible for whatever the claim it is, whether it's money in your account that you're transferring to someone else, or it's a piece of physical property that you own. And so, really, the way we look at it is verifiability and transactability. And so now when you look at to round up, you look at the developments on public blockchains, what they're unlocking. And if you look at a simple example, and I'm looking at it from a technology perspective, really, how could you access the US dollar before? Either I give you a physical dollar bill, or when it got digitized, you needed access to the US Fed's ledger, so you had JP Morgan, et cetera, intermediating that in a correspondent banking chain. And we know that correspondent bank network has been shrinking across the world because of de-risking, which is a big problem for a lot of countries and individuals who rely on that. And now you have stable coins where anyone with an internet connection gets access to the US dollar. It's actually mind-blowing when you think about it. And now obviously, legal frameworks are coming in. You have the Genius Act in the US, Soon Clarity Act, and you know, there may be also economic risks when dealing with these forms of private money. But fundamentally, what I find very interesting about the tokenized world is you go back to that individual agency. I wanted access to the US dollar, I get access to it, right? Um I wanted access to my own property, the ability to transact it, I have the ability to do that. Um that's really how I look at it. Or, Julie, to your point, some of the technical underpinnings that are allowing this to happen. But broadly, I think what we'll see for mass adoption is um developments in newer kinds of legal frameworks. Because I think a lot of this will happen. It will happen in smaller pilots, few corridors, and so on. But really, large-scale adoption uh will now be dependent on legislative frameworks, frameworks, and clarity to come in, which we're starting to see, but I think it's a it's a longer journey.

Julie Verhage-Greenberg

And how does this play? And I remember one of the things with FedNow when it first came out is you know, there's a big difference between like sending the money and receiving the money. And a lot of banks signed up to do the one, but didn't want to sign up to do the other because of the risks involved in the money not arriving, etc. How does it impact it by thinking it, thinking about it in a different way with the tokenization where it's transferring the claim of it versus transferring the money itself, if it has any sort of impact on the and where that kind of uh stands, Michael? Again, views expressed by you and not the Fed.

Adoption Hurdles And User Safety

Michael Lee

I want to take a step back, actually, and I think it's really careful. Uh, it's important to be careful about what the innovation is. I'll give you an example. I I mentioned earlier that we need to be careful about what we're talking about with respect to the feature of money and the way in which it's settled. Like stable coins, for example, are accessible to all, and that has fundamentally changed how we allow for money to flow. There are also still claims to the issuer. And so we still need to be careful about what we are talking about with respect to the power that the underlying settlement infrastructure allows us to achieve and the actual asset that we're thinking about. When we're thinking about a debt instrument that's trading in a tokenized system, that's still require, that's still a form of claim with derivative rights and other types of protections that are afforded to the holders. And so a lot of the innovation that has to happen is not just technical. It's legal, it's regulatory. It's understanding that the limitations are not that we don't have the technology necessary in order to have these transfers occur. It's that we have these types of safeguards in the traditional system that may impede in terms of access. Uh, if any bank in the US could create an account for any individual, then the dollar would be much more accessible to the world. But we have good reason to also safeguard that and to have some compliance functions in order to ensure that the banks are being responsible and therefore being good stewards of the financial system. So I think in that respect, you know, what is really new? Well, there's huge and fascinating uh technologies and arrangements that are possible in a blockchain-based system, for example. And I think, you know, understanding what's new about that is is going to be the key. One thing that excites me, for example, is privacy preserving technologies. We're able to now guard the data from everyone if we wanted to. And that changes the landscape for the types of technological commitments that we can have towards privacy.

Julie Verhage-Greenberg

Can you want to check that a little bit for someone that's more payments 101 than payments 801?

Rapid Fire And Closing Thanks

Michael Lee

Definitely. So we can now devise systems where none of the participants, except for the people privy to the transaction, and also not even the operator has the power to look at the contents of a transaction. So that's the the extreme case of privacy that we can achieve. And we already have systems that are accomplishing this extraordinary feat. So we have the technology to do that. Now the question is, how do we find the precise balance between providing individuals with that technology in order for them to have hard form of privacy that we've never been able to imagine before? And the next step is how do we also accomplish compliance in this framework? Where is the boundary of how privacy can be can be provided as a service from any private or public payment system provider that allows you to actually also be compliant and make sure you're safeguarding the system from really bad actors that we know would enjoy the anonymity of this type of system. So some of my uh recent work kind of delves into this question because we now have the technology to commit to a provision of tech of privacy that allows uh only individuals or agencies or governments to look at the contents of the payments when they have fit certain types of conditions. And that is kind of the basis for what we think of as really importing to the technology a form of trust that we weren't able to bring in before.

Kabir Kumar

That's uh I think there are two sets of things, if I'm hearing you correctly, Michael. One is you're saying, look, technologically there are certain things that are possible now that were not possible before. And the question really is the sort of governance, regulatory, design choice decisions that we need to make. Because let's take with privacy. In the past, we would try to achieve that through a set of sort of laws rather than what was possible technologically. And now you're saying technologically we are further ahead on that, and now we have to sort of figure out what those laws are. There's another question which I found uh in your paper on uh sort of optimal design of tokenized systems, if I'm remembering it correctly. I think that there's this question of sort of what the new technologies offer, uh, what is inspiring the vision of the FINTANET effectively, and what's possible with the traditional systems. So how how do the two coexist? Could you say a bit about that piece as well when you think about the you know the optimal design of these systems? What what the what works well hand in hand with the new technologies and the existing systems that we have today? And in the payments 101, payments in the payments 101 language.

Michael Lee

Thank you. Well, so I want to come back to a point that Siddharth mentioned that I think uh resonates strongly, which is when we think about the new types of arrangements, uh it comes the boils down to having really precise verifiability of ownership. And in in our work uh with co-authors, uh what we envision is a future where we can verifiably demonstrate ownership over not just existing assets of today, but of the future. So a payments one-on-one example would be: are there ways in which we can commit future cash flows after an individual? It could be salaries, it could be income in a way that allows for me to unlock different types of credit arrangements that I would not be able to do because the the cost associated with making or fulfilling this arrangement would be too high from today's standpoint. And I think this brings us down to the question of well, who is able to actually do that today? And this is where enterprise and wholesale uh users of our financial system have access to. And so a different way to put this is by having a trusted environment where we can now have verifiability, we're able to have individuals also have stronger forms of commitment and verifiability, allowing them to unlock their assets or their ownership in a way that we weren't able to before. It's a form of democratization of these financial tools that are afforded to larger institutions or firms that enjoy that because uh financial institutions and intermediaries have the incentives to provide these bespoke products for them.

Julie Verhage-Greenberg

Well, one more one more thing I want to dive into because we we've mentioned a lot of buzzwords at this point, but we haven't mentioned a buzzword, which is Gen AI. And I want to have each of you answer sort of how you think Gen AI will play a role in money movement in the future as well. Maybe Siddhar, we'll go back to you.

Siddharth Shetty

Yeah, so I feel um uh AI plays a very critical role because if we imagine today a lot of our payment experience, so there's two parts there's the wholesale payment experience and the retail payment experience, right? So um, keeping aside, let's say we we focus on the retail side for a minute, a large part of the retail payment experience today is fundamentally driven through traditional UIs and user interfaces, right? And it's that's very fragmented. In some cases, you have interoperability implemented, but you know, by and large, if you imagine cross-border flows, you imagine even domestic flows, you're often like, okay, what app does the merchant have? What app do I have? You're like discussing, do you have this app? Then we start to pay with each other, right? And then depending on how much you want to transfer, then you have to figure out do I use my bank app and so on and so forth. So there's a lot of fragmentation that happens in the user interface for these kinds of payments. So now if you imagine a Gen AI or really an agentic world, what you're now going to unlock is one, a whole new kind of experiences. You know, why do I need to check out? We are so used to checkout pages, put your card details. That is irrelevant. In an agentic world, I of course need a safe way to link my wallet, ensure limits, and ensure the agent doesn't go rogue so that it's overspending my money. And people will build appropriate dispute resolution systems, observability and tracking and so on. But fundamentally, once I link my wallet, that gives the agent agency to pay other agents or other individuals or businesses on my behalf. And so I think fundamentally it changes the transaction medium around how people experience payments, how people not just experience payments, but also the choice. I'm really limited by the effort I put in. Do I use, you know, wise, do I use my bank? Do I use a crypto rail? Even within crypto, which currency do I use, which pair to figure out which is the best rate in the fastest possible time and which one is my beneficiary on. So today I'm I'm really restricted by the time and effort I can spend. If I delegate through this to an agent, the agent can fundamentally do a search and discovery of what are the best liquidity options in the time period and cost structures I want and go ahead and do that transaction. And so I think there's a lot of interesting both on the experience side, both in terms of also the choice and the cost structures. I think there's going to be a deep impact that agents will have to payment experiences going forward. And again, that's fundamentally where tokenization comes in. Because at the end of the day, if we truly want to unlock value that agents bring to society, you need to tokenize underlying assets because otherwise you can't execute on it. I can have an agent, I can have an agent create a will for me. But if the underlying assets are in paper, I'll still have to go back and update paper records, right? And so agents and tokenization sort of go hand in hand where agents can drive a lot of the discoverability and so on. But if you create natively programmable assets, whether it's money or other types of assets, then you can also drive end execution of those transactions programmatically.

Kabir Kumar

I think that's what you apply. Because I can have a digital certificate, you know, but tokenized, programmable, that's the next layer here that you are envisioning.

Siddharth Shetty

Yeah, so there's a key nuance in this, as you as you rightfully brought out. You know, you've gone from paper records to fundamentally digitized records, um, and then from digitized records to tokenized. And so the real question is what is the difference, right? Both seem to be in a digital form. What is really different? Uh, you know, is one just more digital? Uh but the real nuance that that I realized through this journey was fundamentally the universality of access and control of that asset, right? So let's take an example. Let's say you've digitized a security, okay? And now you want, as a retail investor, you want a loan against that security. Uh if you're in the US, you're limited to whether that broker, which is you know holding it, has integrated with a set of lenders to offer you a loan. Uh if you were, let's say, in other countries where it's more centralized away from the brokers into depository infrastructure, you're dependent on did that depository integrate with other lenders, it's a very painful process. If the underlying asset is tokenized, now at a global level, anyone can plug in. It's a simple set of smart contracts on that asset to lock it. And whether it's a security or whether it's representing my house, it doesn't matter. Now, obviously, there is some nuance to this as well, which is depending on the type of asset, depending on the type of jurisdiction, you will have flow control. So you will have capital control rules, you may have, you know, a whole bunch of other rules on who can be an investor, who can be a lender, et cetera. And those can be programmatically applied, programmatically supervised. So those are things that are layers that will get built on top. But fundamentally, the way you want to think about tokenization is it unlocks universality of access and control, which in the traditional digitized forms essentially come down to I have to now do all these API integrations and do all these bilateral contracts. So that's sort of the subtle shift, I would say, between the two. And then, of course, there are still evolution that needs to happen because even if I've tokenized the asset, um, you need some form of legal recognition in many cases for counterparties to accept that form, just like it exists in the digital world. Otherwise, what happens is financial institutions will say, give me a paper record, because my regulations still say paper. You know, even though it's tokenized and cryptographically verified and better, they'll they'll fall back to paper.

Kabir Kumar

And I think the link to AI and agentic commerce, I'd love to get your reaction to this, Michael, is that you you can get higher degrees of autonomy if you are transacting on top of a tokenized universe versus just a digitized universe. I think that's uh one sense we have is that you know you can see a greater degree of truly agentic behavior because you're now transacting with tokenized tokens and tokenized systems. I don't know what do you think, Michael, in terms of these worlds coming together, uh, just to make sure we round out, as Julie said, all the buzzwords of the day.

Michael Lee

So I want to begin by emphasizing the journey of tokenization is just starting. And so while we aspire to imagine how the future will look like, uh, we still have really core challenging problems to solve just from moving to a tokenized system or having at least a tokenized system that is interoperable and allows for us to transition into a more efficient system. Uh having said that, I think there's two things that we need to think about if we are reimagining uh a world where this form of infrastructure and technology is used in mass. And I think this uh insight kind of applies both to tokenization and to agent tech AI. And we have to understand the hurdles required in order for mass adoption to happen. From a consumer standpoint, what works works, and the the headspace allocated to understanding and using these technologies is very, very limited. These things have to work and fundamentally improve as a product or a service the lives of individuals in order for it to be adopted. And it doesn't matter how shiny or new it is. And I think this principle is the same not just for tokenization or for agentic AI or more broadly, any kind of fintech. It needs to solve a problem for individuals and it needs to solve it in a way that's safe, that provides an enhanced experience, but also allows for them to lay to engage in this without facing huge risks. And I think this is where the balance arises with respect to agentic AI. We understand the potential of it because of how powerful it can be in terms of individuals being able to delegate cumbersome, uh complex tasks, uh, allow to streamline a particular transaction or to uh manage their wealth or their finances or their positions in a particular financial transaction. These are all exciting, but at the same time, we need to have in parallel a development of not just compliance, but the controls and the ease of access necessary in order for individuals to be properly educated and then to use these systems in a safe and compliant manner. And I think this is gonna be an important aspect to watch over if we really want to have a system that truly is adopted. Because in the end, if we push for adoption, just looking at the allure of the technology without thinking about the safeguards, without thinking about where the mainstream uh users are, then we're going to end up in a situation where all of the key attributes that we think this can unlock, the inclusion, the access, the governance, all of this will be only used by very sophisticated and already well-endowed individuals who have this access already, who are just using this as another stepping stone.

Julie Verhage-Greenberg

Amazing. Well, I think that's a good place to end the overall conversation, but we like to wrap up these episodes with a nice rapid fire round. Um, so I'll start with Michael and then Siddhar will answer, and then Kavir will ask the next question, and Siddhartha will go first, Michael second, so then it kind of evens things out a little bit too. But uh, Michael, if you could have any dinner guest, past or present, future, who would it be?

Michael Lee

So I'll give you two answers. Um I think a no-brainer answer is Satoshi.

Siddharth Shetty

Uh if I was betting, I knew you would say Satoshi.

Michael Lee

I don't think I need to explain why, but it there was an understanding that having a censorship-resistant system could be could potentially better the lives of everyone. And it started with this very simple and clear principle that allowed us to now uh be in the world that we live in right now. If we take a more historical standpoint, I think one person that comes to mind is Ben Franklin. Uh he was an innovator, but also really pushed hard for uh for paper money to exist, which is kind of the start for where a lot of the monetary innovation happened. And so I think going back to you know the idea of of independence and uh the exercise of freedom and the financial inclusion, he would be also someone that I would want to have a dinner with.

Siddharth Shetty

Siddhar. Awesome. So uh so one of my recommendations, maps, similar to Michael. Um uh, but I'd say first off, uh being very much in the entrepreneur seat, uh a person that I would love to have dinner with would be Musk. Um I know that's a political name, but um but really in terms of you know, how do you build space technology, how do you build brain computer interfaces, how do you build uh automobile company, now a robotics company do all of that at once. So I think there are a lot of learnings there, just in terms of how do you accelerate towards the future that we are going. Um and there are a few interesting ideas we've been discussing internally on what does the multi-planetary version of money transfer and so on also look like without all the latencies and frictions that we have. Um so that would be one, and then the second, uh, very obviously would be Swatoshi as well. Um, um, and particularly from the lens of also the paper talked about how Bitcoin could solve for peer-to-peer payments, and we saw Bitcoin go in a certain direction. Uh, there are, of course, developments now to bring back greater programmability and payments onto Bitcoin with recent efforts, but I'd really love to understand both, you know, just the design principles and how he, she they thought about it, and then um um what sort of would sadden them about the way it's played out, or what excites them about the way it's played out over the last uh few years.

Kabir Kumar

You know, all of us, I think, should have uh dinner with Nandan Nilakani. That would that's something I would add, because I think Nandan is a role model for how billionaires need to behave in the modern era. That would be my two sets. Um, you know, he has envisioned so many amazing sort of public. Yeah, I know. You put Elon before Nandan, you know. I don't know. What is Nandan got to say about this, Siddhartha?

Siddharth Shetty

The only reason I didn't think I'm extremely grateful to have him as a mentor and therefore have to be a good idea.

Kabir Kumar

Yeah, you you have dinner with him all the time. Question for me for both of you. Looking back to last year, you know, any highlight from last year in your work that you think the world needs to know? You know, so we'll start with you, Siddharth, with the Fittenet Labs, any achievements, any progress, any challenges perhaps that you faced last year that you think are worth highlighting? And then Michael would love to hear.

Siddharth Shetty

Yeah, so I I would say the big thing for us has been the establishment of the lab network. So we have one in the US, one in Singapore, one in India, and one in Switzerland. Um, and I think that's really giving us a lot of the input. So we spent most of last year with public authorities and private sector around the world, really learning about, you know, how do things work, what are the concerns, what are the stress factors, and so on. Um, and I think what we're really excited about getting into this year is just shipping, uh shipping new protocols, shipping a whole bunch of other infrastructure and see how that ships up. So a lot of challenges as part of that journey. Maybe that's a separate podcast altogether, but um but I'm really excited about you know the lab network that's set up and the people that have come together and institutions that have come together to support it.

Michael Lee

On your end, Michael? You know, I I I think it, you know, 2025 has been an eventful year, even maybe historic. But I I what I really think has been a highlight is the types of conversations that I've had with a very diverse set of stakeholders coming from the technology side, from the blockchain and crypto uh sector, all the way up to institutional investors and players and regulators. Uh, we've seen a confluence of uh discussion that has emerged. And being part of that and understanding the perspectives of this very broad set of constituents who care about uh building the next uh financial infrastructure, it has been a rewarding, enlightening, and uh incredibly grateful experience for me.

Julie Verhage-Greenberg

Amazing. What one quick one I'll ask then and keep it a short answer. How do you de-stress? It could be exercise, cooking, cleaning, sleeping, tokenizing, tokenizing. Sleep hits different with a toddler. I I I concur on that one. Siddhar, what about you?

Siddharth Shetty

Yeah, yeah. Sleep also hits a bit different with cross-border travel. Uh but uh but I would choose uh I I would choose uh movement. Uh I feel movement for me is is a great way to just de-stress and refresh. And and particularly it could be a combination of either running or I grew up by the ocean, so big fan of swimming where possible. Um uh but I would say it's really a combination of these two that uh uh that just get me back into you know a much more grounded state of things.

Julie Verhage-Greenberg

Amazing. Well, thank you so much, you guys. I really appreciate the time zone and stuff too. Like it was you tell it was dark when we started this podcast out here. Sidar, it's probably dark where you are now. Um the the time change this weekend kind of made it so you had to do it later since you know the US still does daylight savings and other areas of the world do not. Uh, but I really appreciate you guys taking the time and giving me a payments 101 lesson on tokenization and all the fun buzzwords around it.

Kabir Kumar

Really enjoyed it, guys. Really enjoyed it. And have you both of you met before, Siddharth and Michael?

Siddharth Shetty

Have you no, we haven't. I think you guys need to spend more time together. We gotta we gotta we gotta figure that out.

Julie Verhage-Greenberg

Yeah, we did just them and Satoshi at dinner, and you know, we're good.

Siddharth Shetty

Forget Satoshi, you know, we just need these two.

Julie Verhage-Greenberg

Perfect.

Kabir Kumar

Thank you for making time for this, Michael, and thank you, Siddharth, for making time late at night. And I genuinely believe the two of you should interact. And Julie, thank you so much for doing this with us. Really appreciate it.

Julie Verhage-Greenberg

Thank you so much for all the hard work behind the scenes as well.