FEDERAL RESERVE: Future of Payments & Interoperability | J.P. Morgan, Chainlink, Amazon

I'm Josh Lipsky. I'm the chair of international economics at the Atlantic Council. Joining me to my left is Nlesh Dusain, global head of institutional payments at AWS. And as Jula said on the Zoom, we have Umar Farooq, co-head of global payments, JP Morgan, and Sergey Nazerov, co-founder of Chainlink. We are going to keep this dynamic, free flowing across multiple continents and time zones. So, Umar Sergey, thank you for doing this. And Alles, I'm going to start with you and I want to pick up on something Governor Waller because I thought that really frames so well these issues that we're talking about in this session about embracing a multi-chain world with crosschain bridges. Long title, but really what we're talking about is technological developments around the world, the future of payments and what's changing. And I want to set the scene a little bit. The title of this session is embracing. Is there any choice but to embrace what's coming?

>> Sure. No thank you so much for uh having us here. Um so the way we uh we see this right which is we are already uh just to do a level set um when we do payments or when we you know send payments whether it's domestically or crossber uh we do we make those payments within a certain network uh so for example you know we use our banks or bank bank accounts to make payments uh we like uh using for some use cases uh solutions like Venmo or PayPal. So those are all networks and the new networks that are coming on board like the digital asset based or tokenized networks those are still like you know depending on what the customers choose uh they'll use those networks for their appropriate use appropriate use cases. So in some sense from a payment standpoint we are already used to living in a multi-et network environment and the question is okay based on the new networks that are coming out which is you know based on the digital assets or other networks what are some of the use cases like governor Wer said the private you know private sector can come up with and how can customers whether they are uh individuals like us over here uh businesses small to mediumsiz businesses or corporates can use those networks for various uh payments use cases. And that's where we are at right now which is okay how do we um uh you know take the learnings from all the decades of you know operating in a multi-et network environment when it comes to payments to using these new networks that are coming up and then using them for payments as well.

>> Thank you. I think that's helpful way to think about it and very happy to have your perspective and private sector's perspective here. This is only going to work public private collaboration going forward. Umar I want to turn to you. You're in Singapore. I saw earlier today Monetary Authority of Singapore announced a new wholesale CBDC project and I wanted to ask you about the digital asset ecosystem we're all going to live in in the years ahead. We're going to have some countries doing retail CBDC, some doing wholesale CBDC, certainly JP Morgan tokenized deposits, stable coins as we talked about. What does this all look like from a crossborder landscape going forward from your perspective?

So first of all uh great to be here despite the fact that it's quite late in Singapore but you know that's the advantage of being heavily jetlagged so you know you don't know what's day and night but uh but I think jokes aside um I really think I think there's a couple of dimensions to think about uh when you when you really approach this problem one is this separation of sort of instrument from infrastructure and you know you talked about infrastructures already and we all have to operate across many of them even without blockchain let alone blockchain world so I I mean it's not like it's something new for us. Um whether those infrastructures are bounded by national boundaries, whether they are bounded by high value versus low value. I mean there's all kinds of payments infrastructures. I think when it comes to things like CBDC's and tokenized deposits and stable coins and I think everyone has many people have many views on many of these things. So you know I I tend to be a little bit more skeptical towards the CBDC topic but you know other people are very pro it uh including Singapore and uh EU but when you really look at it it's essentially the recreation of the current structure in a different technology with different capabilities. So in the current structure and I've said this before in uh other forums you know we have we have central bank money. So we have the entire M0 layer. On top of that, you build an M1 layer which is all of the commercial banks and then on top of that there's various you know M2 plus type u um monies which are private money. So in the regular world that would be you know reserves of the Fed, JP Morgan deposits and then um let's say you know cash in your um you know square wallet or something like that. Um, and if you really think about it, I think this is all basically getting transformed in a digital age into potentially CBDC's, although you could also do that with real-time RTGS systems. Um, you basically on top of that would have M1 layers, which would be deposit tokens, which is something we launched earlier today, which is JPMcoin on a blockchain on a public blockchain. And then on top of that, you have stable coins that will float around as well. So you both have the money stack and then the infrastructure sort of what I would call the horizontal that's as as you talked about is you know somewhat fragmented right now and maybe it's going to be lots of bridges and maybe it'll be fewer bridges with fewer uh infrastructure surviving long term.

>> Let me just ask one quick followup and then Sergey I'm going to come to you. Break out for us a little the retail or wholesale. We talked earlier today with Governor Waller about crossber payments and stable coin usage. We haven't seen stable coins expand yet on the wholesale side at the scale that you would want to see. Obviously, that can happen. The infrastructure is building up. How do you see that developing? Is that a role tokenized deposits will play? You might have a dog in that fight.

>> Yeah, I mean we definitely have a significant dog. It's probably the only one with a from a large bank at this point in the public chain. But when you think about wholesale, some of the same things that might appeal to retail do not appeal to wholesale clients. So for instance, you know, it's not like you're going to have a whole lot of wholesale clients who are not banked and they don't have bank accounts already. Uh so you're not going to be in those economies. A lot of wholesale clients are going to operate in larger economies whether that's the US or India or you know parts of Middle East or China etc. So you know those are actually relatively efficient corridors even for retail but definitely for wholesale. So when you think about wholesale money movement, the things that are important to clients are obviously the efficiency of movement, the certainty of where their payment is. And frankly, somewhat most importantly, they don't want fragmented liquidity. So they don't even like the fact that they have three big banks, you know, let alone that they would have banks and then they would have stable coin and now you're, you know, tied up into an instrument that potentially is paying no yield or at least no explicit yield. And suddenly, you know, you have all of this complexity to manage. So for wholesale clients, they do want the upside of blockchain tech, possibly public blockchain technology, but they don't really care about anonymity. They don't they're not really unbanked and things of that nature. So that's why I think the take up has been slow. And what we are seeing is a lot of clients have come to us and said for wholesale money movement both cross borders and frankly also on public chains. They'd much rather have a deposit token where they have a certainty of what the instrument is. They can get yield and they don't feel like the liquidity is actually at that point fragmented because all the liquidity is sitting in a deposit account whether in the shape of a deposit token or a real DDA. Retail however is a slightly different story and I know we'll probably come back to it a couple of times. Um the the obviously the pitch here is you can go from point A to point B lowost instant. Um I would argue that there are corridors like let's say US India, US Philippines etc. where there's lots of big players and not necessarily banks but fintex like the likes of wise and remittly and others that actually do provide a pretty good experience if you have a bank on the other side which a lot of these uh you know sort of remittance acceptors do. However, when you get to geographies with a high proportion of unbanked clients, high proportion of people without proper KYC for instance, not that they're bad guys, they just don't have ID etc. In those cases, um, stable coins do, uh, frankly provide a service that, uh, general existing technology probably cannot.

>> Okay. Well, you led me right where I want to go, Sergey, to you, which is Umar talked about recreating the system that already exists with new technology. There's another part of the system that has to be recreated on compliance, on AML KYC, on the way that we monitor these transactions around the world and ensure they are safe. How does the new technology allow us to do that and where are the gaps right now that you see?

>> Yeah, sure. So all that can be automated and you can basically create onchain conditions and smart contracts which we already have with something called ACE the automated compliance engine and that's a combination of contracts that define compliance conditions and identity and a set of off-chain oracles that feed the identity information in and can synchronize with policies that already exist. So you you're going to need to operate those policies on a specific chain and then you're going to have to operate those policies across chains and then the complexity it progressively grows because you know Umar for example is in Singapore and I'm in the US. So now we need to meet two sets of policies right we need to meet Singapore's policies we need to meet US policies. So the transaction for example might originate in a US chain and meet the policies of you know the US stable coin issuer or the US system for how stable coins are allowed to move into Singapore and then there's a smart contract that the coin is controlled by that doesn't allow it to move unless certain things are verified onchain certain conditions of identity like not being on the OFAC list other conditions Right. But then you have a cross-chain bridge from the US chain to the Singapore chain and now you need to meet the conditions of the Singapore chain and now there needs to be a contract and a and actually the bridge needs to meet those conditions. So that's where something called CCIP comes in the crosschain interoperability protocol which is like the TCP IP of the internet. So the bridge also checks those conditions and then when that stable coin or tokenized deposit or CBDC, you know, and I'm a I'm a big believer in what Umar is talking about with tokenized deposits and really there's going to be different groups that generate different tokens to meet their uh goals, right? So tokenized deposits will meet a lot of user group goals. CBDC's for others, stable coins for others, which you know Umar was right in creating that starting to create that bifurcation between groups. But my view is that all of them will will operate against a set of onchain conditions in both of those chains, right? Both the US chain and the Singapore chain. And then the bridge will also need to verify some conditions. But but the good news about this is that those conditions don't need to be housed privately inside of the servers of one of the counterparties such that the regulator it can't get them easily or it can't be verified easily that compliance has happened. In fact, the verification of compliance will go way down and the cost of compliance overall will go way down, which I know is really funny to say because the the blockchain industry is not not well known for its tendency to do compliance, but you know, that's that's kind of like a different part of the blockchain industry. I I actually think that onchain automated compliance will create huge efficiencies for the um for the compliance side of these payments processes processes and it'll I I think the regulators should actually create clarity about what that onchain compliance looks like because that's how their um transaction monitoring will work, right? their transaction monitoring will work against that onchain compliance which is what we're working on with with multiple regulators including the MAS and and and other and you know we've had meetings with the SEC on this topic and the Treasury and and other groups in in in the US.

Just one followup on that because you spoke at the Fed last month, you talked about your work with Swift and I was wondering when you spoke about that. Are the legacy systems here, Swift chips, the way we do crossber messaging and payments fit for purpose for this new world we're moving into, can they adapt to the compliance regime you're talking about or is it going to be a yes and situation of others moving into that crossber interoperability space?

So the the systems like Swift and AC and others uh have a few key components that are reusable. Some of their components are non-preusable. The components that are reusable are the private keys. So all the devices and software that Swift members and AC users utilize to sign transactions cryptographically. Those keys in that software can also sign blockchain transactions. Also there are shared standards. The ISO 2022 standard is is very widely adopted by you know the Fed as well as Swift. So within that standard you can make payloads about blockchain events which is what we've been doing with Swift. We've been making payloads that in a ISO 2022 compliant message say move this stable coin from blockchain A to blockchain B. And that payload is signed by the same key that would sign a Swift message to instruct it to do a traditional, you know, payment. What what isn't reusable possibly is the routing infrastructure, right? So if you look at what what these systems are, they're signing keys. They're essentially a definition and data dictionary and like standard for defining transactions and then their routing infrastructure to do final settlement, right? the the routing infrastructure is what's really getting replaced by chains and bridges, but the private key signing infrastructure is completely reusable. We've proven that with Swift over many years. That includes H AC related signing keys. That includes Swift signing keys. All those keys and are reusable. The payloads that they sign need slight modification, but it's not a complicated process. We've done it multiple times now. And um that we think and I think is actually the most efficient way for existing systems to interface with blockchains. It's for them to use their existing key infrastructure and their existing like ISO 2022 standards, but for those payloads to say things like take a stable coin, move it from the US blockchain under these compliance conditions over to the Singapore blockchain and you know return a confirmation back to me also an ISO 2022 message. So the answer is a lot of it is very reusable. I would say the vast majority of it is, but the final routing and settlement is what blockchains and bridges are replacing. I I just want to underscore this because we are appropriately wonking out as we should at the Philly Fed conference and it's very technical and it's incredibly important but there's a geopolitical lens on this because if you think about the way the US government and G7 leverages the swift system for example on sanctions so it's it's not just about the crossber settlement it's about foreign policy national security geopolitics which is also going to change we don't have Tyler here to ask him that question but that will be the followup going forward Nlesh, I want to come back to you and talk about the G20. US has the G20 presidency coming up in a few weeks when South Africa finishes. Then the UK has the G20 presidency. Now, the crossber payments targets are not going to be met. The FSB said this. They said this last month. So, we need a reset. We need a reset in the crossber payments targets. We're not meeting the lower costs. we're not meeting the speed which is kind of incredible because if you think of what's happened since those targets were set four years ago we really should be doing much better why are we falling short and what is a reset of the crossber payments target look like

>> sure so u I'd like to pick up where Sergey left which is if you think about any payment process there are three aspects to a payment process one is like the messaging system right which is you know in the financial industry ISO 2022 has been becoming very um uh you know adopted very widely And while it is being adopted very widely, it is especially in the context of crossber payments. Uh there is no like one messaging standard which is which goes across different networks. So for instance uh between two bank accounts, sure you know you could use swift, you could use ISO 20 or 22 messaging standards to you know pass the uh payment information back and forth. But if the beneficiary on the other side is not does not have a bank account and they want to take get the payment in some from some of a digital wallet then the payment network that is operating that digital wallet needs to be on that same messaging standard. So one thing that um I think should happen at a at an industry level and AWS is not like in that area right which is the industry should come together and have across payment network some kind of a common messaging standard. So that's number one. Um, number two is the movement of assets and that's where I think what Omar was talking about which is whether it's preunded in a bank account or some form of a digital asset or a tokenized asset going back and forth across different different chains that becomes a pre that becomes a follow-up to the messaging part right so say you send a message that someone from the US is sending money to someone in Singapore uh if everything checks out from a standpoint of okay the beneficiary information is right uh all the information around a payment Like if it's a business payment, purpose of payment, things like that, all those things are included in that message. Then the actual movement of money can happen, right? So that's the that's the second part of it. And the third part is are things that happen around a payment process which is around compliance. So you have to do sanction screening, money laundering, all those checks. So for us to as an industry to reduce the what what I think about friction when it comes to crossber payments in different areas and there are certain corridors like Omar said where there is very little friction but where there are other corridors where there might be a little bit of friction or a lot of friction in terms of slowing payments or high cost of payments. I think we need to align on these three areas which is common messaging standards uh ensuring the other party knows what is expected from them before the actual payment is sent then how are the assets going to get transferred whether it's through stable coins or fiat or through other types of uh tokenized uh deposits and then finally uh the the compliance checks right and we need to work with all the entities who are sending these uh payments before they are actually sending the payments to do all the checks that are necessary to to so that the payment is successful and not get stuck in like a false positive or something some process where you're not able to get um send a payment to the beneficiary. So I think from a G20 standpoint if there is alignment at those three levels I think that'll be that will be helpful to drive drive drive the cost down and then reduce the friction for crosswater payments. I I think that's helpful to get away from some of these specific number benchmarks which obviously we're not going to hit versus what do we want from a system interoperability ecosystem digital assets crossber payments that's healthier and focus on the rails which I'm hearing from everyone not just today but going forward is kind of the theme of these conversations that I've been a part of

>> Umar I'm going to come back to you let's talk about interoperability I mean Sergey talked about it chain link obviously plays a role here but how do you see interoperability going forward is there a role for the Fed? How is this going to work in this digital asset ecosystem you outlined with the traditional players like JP Morgan and the new players in the system?

>> So again I I would look at interoperability across instruments or across different monies and across you know sort of in infrastructure which actually is where Sergey's team plays a very critical role and they are developing solutions. But if you think about let's say let's talk about like tokenized deposits or deposit tokens and you can have a JP Morgan dollar and you can have a Bank of America dollar and city dollar and Welss dollar. So you have you know all these dollars floating around. I think one thing that again I'm sure everyone on this call realizes but actually generally people do not realize is that uh every bank's money is slightly different from every other bank's money despite the fact that they are dollars and the Fed actually plays ultimately the role of creating the singleness of money across those institutions in some ways because they're all essentially at the back end is the Fed is the sort of you know the underlying foundation for all their monies um and I think you need something similar to happen in the new digital age as you want to basically maintain the singleness of money to some extent and the way we do it right now is we will have a deposit token but then if you really want to get to city you would have to then move it back to you know regular dollars going out to city it's not going to be the deposit token going to city because what does it mean for a city client to hold a JP Morgan deposit you know token so to speak I think it's a starts getting into some really complex stuff that obviously those of us who are like sort of payment nerds really you know love to discuss but in reality most people don't want to see. So I think that's this there's that sort of singleness. There's singleness across stable coins where you know there's such as a result of genius. I think you see like everyone and their uncle trying to launch a stable coin and if that's the trend that continues you're going to end up with a lot of stable coins looking more like loyalty points than money because then you know what do you do? I mean you show up to a coffee shop and then you figure out which dollar you're going to use there today. I mean that sounds like you know having airline miles. So, so I think I I think there's a singleness role that you can play. I think when it comes to infrastructure that is frankly harder for public institutions to play. I think that's where you know companies like Sergeys and others would probably play a bigger role from a technological point of view. I would say I am frankly a little bit surprised by direction of travel. So we've been in the space for about 10 years. probably have the most uh the deepest and the biggest uh team in the space of all large banks. Um and you know we've been over the years uh accused not once but many times of having world gardens and having our own blockchain and stuff of that nature and you know how public chains are the only thing that's ever going to happen and then all of a sudden now in the last last year or so every large entity is launching their own chain. So you just go okay that's a bit bizarre. So I think that's you know it's it's probably going to go from if you asked me three years ago I would have said oh there'd be a few public chains and they' basically rule everything and now it's a bit unclear how many L1s and L2s are going to exist in the short term, medium-term, long term. So I think that's also um sort of going on. I I you know I think the one thing I I probably would come back to because you asked this question about the G20 targets you know uh again the numbers are the numbers and I do think for many of the corridors as we discussed I mean if you go with a 3% target on average I would say some of the ma some of the major corridors are probably operating sub 3% for you know a lot of the money movement um and this is where the fintexs come in in reality they are charging a pretty thin spread on FX. So, it's not like they're charging 5%. That number starts to go up the less liquid the currencies are and the less bananked the clients are. And while stable coins can solve some of that problem, one thing which wasn't coined by me but by the CEO of another fintech that is in the money movement space, you know, you have to look at the stable coin sandwich in the current world. And what that means is you start from fiat, you go to stable coin, then you end in fiat because the person receiving the money in some I don't know part of Asia or subsaran Africa has to go buy bread. And at least for now, it's kind of hard to buy bread with stable coin. Maybe that will be true one day. But you know, and everything will get dollarized. But you leaving that aside, you would have to go buy bread and they need to move it back to cash. So the end toend cost of that transaction may also not meet the goals of 3% average. So again I I just wanted to sort of you know um address that question as well because it's something um you know many of us have discussed quite frequently amongst banks and stable coin issuers.

>> Okay, you gave us a lot to think about there also on the airline miles. I have to think about the analogy. One quick followup for you. uh you know when we think about what's happening here in terms of the fragmentation issue in the genius act you mentioned this a little bit what do you see actually happening on the ground is genius act raising the bar there's announcements yesterday we were talking about in Brazil do you see the reciprocity happening that I think was part of the intent of genius of others trying to meet those standards or is it too early to tell globally

>> I believe it's early to tell but signs are that people are going to at least uplift their standards at least what I call uh countries that are a little bit forwardleaning. So there's sort of you know three or four groups of countries. There's countries that are forwardleaning and they don't have much but they will probably use genius and ma to be honest like you know some combination of the two to figure out what the right standards are. There's countries like China that have you know taken a bit more of a reluctant stand when it comes to stable coins. I would put India in the same category the RBI. I mean again they have their reasons and I can sort of sympathize with uh folks who want to control their own monetary policy and probably are a little bit you know reluctant to go uh uh you know deep dive into something that may have a different end than they would like. Um so there's those count and then there's obviously a bunch of countries that are just not forwardleaning. Maybe they are smaller economies. Maybe they're not that advanced and they will likely take on some construct like genius and adopt it you know more unchanged but that's more speculation on my part but I do think early signs seem to be that folks who don't have rigs which is most of the world are utilizing genius as a road map frankly and uh Mika as well

>> well that's another good task for the US G20 presidency next year to take genius forward through that group and involving China as well we'll see what comes of that surge I want to ask you a question. This was the number one question when we talked about this session and everyone wanted me to ask you. It's about of course the fat finger incident that happened last month. So 300 trillion minted multiples of the size of the global economy. For those who aren't as well-versed, could you explain how that can happen? How that even though quickly reversed, how something can get minted with obviously nothing backing it at that point. Walk people through the process. This was the number one thing I was told to ask you and I didn't want to leave the session without doing it.

>> Sure. Sure. So I I think the reason you're being told told to ask me is because we're the largest pro chain link is the largest provider of something called proof of reserves which would have uh immediately stopped this incident from happening. So when you think about smart contracts you have to ask yourself also what is the other system that controls the smart contract. In most cases, it's a private key, which is the thing I was talking about before, which is the thing that can sign and instruct the contract to do things. In this case, someone instructed the contract to mint 300 trillion US dollars of stable coin and there was nothing else to check that. So, there was no other system to check whether that instruction from the private key was correct or incorrect. What proof of reserves does um is it puts a piece of code into the stable coin code called secure mint. That's what we call the function that chain link provides. And the secure mint function has a corresponding off-chain system. So the securement code is smart contract code that approves or denies a command. And then the off-chain system, the off-chain part of proof of reserves is something called an oracle which checks the bank account or the custodian API of the stable coin issuer. And so if proof of reserves was enabled with securement in the contract when the instruction came to generate the $300 trillion in stable coins, the next step in the smart contracts logic would have been to execute the securement function and therefore check the actual holdings of the stable coin issuer. And I'm assuming those holdings wouldn't be $300 trillion. And since they're $300 trillion dollars, the smart contract would negate that command. And so this is what I was talking about before about the automation of these compliance and other things. In this case, what you're automating is a hard check of the underlying real world assets in the real world for any minting event, any coin creation event of a stable coin. and and that's what proof of reserves um does. So that incident was completely avoidable and we've been talking at length with multiple regulators, the Treasury and and others about, you know, how proof of reserves works when it works correctly. And my opinion is that it works correctly by creating this piece of code that's embedded in the in the code of the stable coin itself. And then that stable coin code is dependent on an oracle to verify things. And then these types of situations become impossible. They they they don't even become probabilistic. They become technically impossible. So even if someone fat fingers it again, the system protects us. You see? See, this is really the beauty of smart contracts and oracles and all this stuff is that you can codify all of these edge case events and all these compliance and all these things you care and worry about into the contract. And so you can frontload the management of all this risk and all these problems. that that's what's going to happen more and more in the financial system and in the payments world is this programmability giving you new features as well as giving you um a lot of protection and risk reduction. Oh, and by the way, the regulator would be able to see that there was a a mistake of a fat finger and that the con, you know, so they would be able to look into that as well and that that mistake would not be hidden in some system somewhere that they're unaware of. So the reg the transparency for regulators and the efficiency with which they can have that information goes way up. And this is the thing we're also you know explaining to a lot of them which once they understand they're very excited about because the their job of monitoring becomes actually a lot easier with blockchains not harder but you know that's larger topic.

>> There's a debate going on in the Q&A that I won't read all of you and we will follow up on surrogate but you've generated a lot of comments with what you said but I just wanted to drill down on that regulator issue. How does that work? Is this real time visibility for the regulators on the blockchain? How are they integrated into those systems? Because this seems like something that's evolving in real time right now. I mean, they have all kinds of transaction monitoring tools hooked into exchanges. You know, there's there's a million of these of these tools that they've bought from the traditional world. There's there's various good companies out there like chain analysis and others that provide basically onchain analytics. I think the amount of onchain information actually I know because chain link is the largest provider of onchain information from the offchain world. We put the most data on all chains in relation to all these things proof of reserves market prices identity data all these things. Um there is going to be much more information appended to transactions on chains than is even available in traditional systems and then there will be software tools like chain analysis where people can access that information or regulators can access that information. Th this does present by the way a significant problem with privacy. So that's that's where this wonderful transparency thing kind of starts to you start to see a different side to it. But you know that that's once again a different topic and and there are solutions to that and chain link has recently released some confidential computation solutions which solves some of those problems. But um so it's not it's not all wonderful right you you want to create the transparency that is easily accessible but to the correct parties. So the counterparties, the regulator, you know, maybe certain select market participants like rating agencies, but you don't want in the traditional world everyone. So this is where you know you have a kind of um fine line to walk which is which is a more complicated topic but for regulators generally this makes surveillance and assessment of transactions cheaper and easier. The the problem is our industry has been colored for many years with like st like bitcoin and dark markets and that has just created a perception of illicit finance. But when you look at the reality of how how the technology works, it it's it massively reduces money laundering and illicit finance. The illicit financing rates on chain are much much lower than the elicit financing rates in the traditional system. they're multiples lower. So, you know, that's the thing that people don't understand because of the historic, you know, negative stuff that has happened with with systems that weren't built that that to do that.

>> That's a key point and I'm glad you said that and it's something we do a lot of work on at the Atlantic Council trying to bring this data to bear because there's a lot of misconceptions in the space about what is and isn't happening. Allesh we've danced around the topic of AI but we cannot have any session anywhere in the world without talking about AI. So basically when I do these events we talk crossber payments and now in the past year everyone wants to talk about AI and I want you to help me bridge the worlds. Sure

>> I can sort of see it but it's helpful to bring it together. How does this actually >> develop into crossber payments if it does at all? No absolutely I yeah we can't have a conversation and not talk about uh AI or generative AI. So uh first like the financial services industry has been using uh like the traditional AI and machine learning for many many years. So that is not something new here. What is new over the last call it couple years is the emergence of generative AI and that's where the real um uh excitement is in terms of trying to figure out different ways of using uh generative AI technologies and when we work with our customers generally uh the different use cases our customers are using generative AI for fall into three buckets. uh one is productivity uh productivity could be for your internal engineers developer productivity productivity or things like that or business process optimization. So in the case in the context of crossber payments uh you know there are several processes customers have to follow to to send these payments more efficiently or to the right uh you know based on the right information. So there are areas where genative AI can be used to optimize some of those processes. That's number one. The second area where our customers are looking at using generative AI is around uh risk management. So how do you identify transactions which are uh not uh in the in the uh they're they don't follow a regular pattern. So say a consumer or a business is sending an average transaction of uh $1,000 a month or something like that and suddenly you see a transaction which is coming up $5,000, $10,000. So technologies like generative AI can proactively identify those transactions but also with appropriate guard rails and checks alert the man in the middle whether it's a payment operations person or a risk management person at a financial institution or a payments company to give them those alerts and say hey there is something you know worth looking into over here and then that person payment operations person or whoever right looks at it if that is really a transaction worth flagging then they can create a message again using generative AI and send that to the customer to say hey is this something uh is this a transaction that you've initiated or is there something going on that is worth looking into. So uh you know risk management and better protection or you know reduction of fraud is the second area where we are seeing genative AI being used uh by our customers quite a bit. And the third which is very exciting is new value creation. So we've talked about u you know hyperpersonalization or personalization in a lot of instances where you want to offer a uh experience which is very personal to an end consumer or to a business or to uh to a large to a large corporate. Now personalization as a concept is not a new concept. You know we've been as an industry trying have tried to uh personalize experiences for many many years. What is different in this case is personalization at scale. So if Josh is looking at his, you know, payment screen versus I'm looking at my payment screen, I with generative technologies, it gives my payments provider or my bank the ability to tweak my experience just so that it's relevant for me. And then the same thing can be applied to uh small to mediumsiz businesses or to corporates as well. So if you're a small to mediumsiz business say in the healthcare space the screens that you might be able to see for initiating payments would be very personalized or you know depending very very targeted to the industry or to the business you are in versus if you're a business in the in the transportation industry. So those are the types of things that generative AI can now uh enable the industry to uh offer very custom or personalized experiences and more importantly depending on uh which part of the world we are talking about embedded experiences. So there are uh regulations in other in in different parts of the world around embedded finance or open banking or things like that. So generative technologies can help build those experiences and also help institutions or help financial services companies comply with those those um regulations as and when they are uh they're applicable.

>> Thank you. Um I want to turn back to you. We have about 5 minutes left. I'm getting the signal here. We've talked conceptually about crosschain bridges. I want to talk about some specific projects that are out there and get your view on what's working and what isn't. I was in Hong Kong back in March. I talked to HKMA about Embridge and their expansion ideas. I have in front of me senior Fed officials. I wanted to get your view of Project Agora, which is how I pronounce it. I know it's not how everyone pronounces it, but that debate's been longstanding. But your view of some of these projects, what's working, what's not from the JP Morgan and the private sector perspective.

>> So, uh, obviously all these projects have different aims, so to speak. I think Embridge we all sort of realize the aim is to diversify away from one central or one sort of global platform that as you mentioned you know when you talked about swift could potentially become problematic for certain parts of the world uh under certain circumstances. So I feel embraing forward it's probably not caught on as much as it would have otherwise. uh also you know the announcement of the bricks currency and those sorts of things. I think some members of the bricks might have made comments without maybe consulting fully with other members of the bricks. So a lot of that stuff has sort of you know thrown some sand in the gears it seems uh of Embridge. Um also you know the I think it's it's pretty it's pretty widely known that there has been an approach to maybe promote ECNY broadly uh outside of u of China. So embbridge could have been a vehicle for that but again seems to be in early stages and moving quite slowly to be honest. I mean in terms of where it was a few years ago and where it is now. So it seems the focus is much more on ECNY within China and getting that bit more robust than anything else. Um Agora I mean at the risk of getting myself in trouble in front of the Fed which obviously you know as uh you know JP Morgan has a very important relationship for us. Um I think I mean I I I jokingly like you know sort of uh came up with something called the reverse medaf law which is uh for those of you who know networks metaf's law is like you know sort of the strength of the network goes up by the square of the nodes um and um you know I always have a reverse meduff law when it comes to banks which is the chances of success of any project is inversely proportional to the square of number of banks involved. So, you know, because there's uh legal departments that freak each other out and then suddenly everything gets like bogged down into nothing. So, again, I will not comment on Agora, but I would leave it at that. Uh, and so that's how I I think if you really want to drive innovation, you have to do it with few players driving their design, fully committed, willing to invest, and then broaden it out. I think it's very hard to do stuff across many countries with tons of banks. I'm not saying it's undoable, but it's tough to do.

>> Sergey, I could tell by your reaction you wanted to jump in.

>> You're the best.

>> Okay, I'll leave it at that. Final round, two minutes. So, I'll give you each 30 45 seconds here. And Sergey, I'm going to start with you. So, we'll go in reverse order and let's end with you. What is one thing you'd like to see in terms of the guard rails or the standards put up around crossber payments going forward whether that's from the private sector generated or from regulators if there's one thing that could really ensure this develops with all the potential we're talking about but develops safely what is one that you'd pinpoint Sergey I'll start with you

>> so the market structure bill in the US is uh is really landmark because when you have payments you need to pay for something And so we need more assets on chain for people to want to pay for them. And the market structure bill I think will accelerate the creation of assets and will accelerate the creation of a larger and larger onchain economy where onchain payments will become increasingly important as the preferred method of payment including for crossborder. So I think legal clarity from the US and the creation of a clear regulatory regime for all kinds of asset issuance that payments will be then used against in delivery versus payments transactions is probably the single biggest thing that governments can do especially the US government because as you know the US government is followed by many other regulators. ICO and all these groups follow heavily what the US does. So I think if that happens, we start to see the forming of of a real onchain economy in the traditional finance part of the world, not just the public chain part of the world and that will massively um feed the payments related volumes because you'll be you'll be doing payments in this exciting new economy and the more that economy grows the more valuable and important payments mechanisms are for that economy. So I want to see that economy grow and this that that will be a big the the the legislation and the regulatory guidance from the US will be massive in my opinion.

>> Thank you Omar.

>> I think uh as Sergey mentioned there's the this technology brings a different kind of transparency maybe some limitations in privacy and synonymous sort of nature. I do think the whether it's regulatory bodies or lawmaking bodies really need to think through through you know maybe uh laws like clarity etc how to operate and you know sort of enforce a standard of BSA AML KYC that utilizes the unique nature of this technology so it doesn't try to just copy paste what's being done in the regular world but at the same time I think if we don't have the right safeguards it's only a matter of time before something bad happens and that could be a severe setback for the entire ecosystem.

>> Thank you, Omar. Thanks for being up late with us Singapore time.

>> And we literally have a minute. So, I'll touch on something we didn't speak today, which is the customer experience. And we talk about at Amazon, what does the end customer experience look like? So, for individuals, for corporates, for mediumsiz businesses. So the more uh guidance or you know innovation we can have from a uh building new experience standpoint I think that'll be really cool to not only use the payment networks that we have today better but also use the payment networks that are coming up in the future because ultimately if customers don't like whatever our industry is building then that that's that's for nothing right so I think indexing more on customer experience is also equally important. Yeah, I'm glad you ended on customer experience. We often talk in these issues about what is the use case, what does the user want, and how do we make a better experience than the one that exists. Nesh, Umar, Sergey, I want to thank you. When Julapa said two would be virtual, I said that's going to be tricky, but not with this group. We can make it happen and you proved us right. So, thank you all and best of luck. Thanks everyone.