Building the Agentic Economy — Circle Keynote | Boston Blockchain Week 2026

Thank you. Thank you. Are we here? All right. Thank you.

Sometime uh this spring a piece of software paid another piece of software 30 cents for a data set. There was no human involved. There was no PO. There was no invoice. And that by itself is not very remarkable.

It's just 30 cents. But what actually did happen after that is that there have been there have been 165 million of what they call microtransactions since then. And what's remarkable is that most people outside of a few in this industry have even noticed. Now, I want to be up front that I am not an AI researcher. Uh I am not an AI maximalist.

I still wrestle with how AI is going to impact the world and what our role is in in a social construct so that the world that AI helps us build is prosperous for us all. But I have come to really realize that perhaps the most important question in AI today isn't necessarily about intelligence. It's about money. What happens when a piece of wallet when a when a piece of software gets access to a wallet and can do something with it? That's really what I That's the question that I want to explore with you today.

But before we go to um the future, I want to talk about a story that that happened here. Not Boston more generally, but in Quincy. 200 years ago in 1826, there was a group of Boston businessmen that wanted to do something pretty audacious. The first thing they wanted to do is they wanted to actually build a monument on Bunker Hill using Quincy granite. But the problem that they had was that there was no way to actually move the granite.

There's no way they can move the rock to create that monument. So, an entrepreneur created what America had never seen before, the railroad. And the only thing that railroad was used to do was to actually move rock 3 miles from here to the Neponset River. And what I love about that story is that dysfunctional railroad was only built for rock moving, rock towing, but it became the bedrock of the account of the American economy for 100 years to come. Obviously, we know what railroads have been used in terms of moving freight and mail and and people.

And that one piece of of innovation created this economy. And Circle, the company I work for, has a similar background. Our our our founder, Jeremy Allaire, started Circle here in Boston in 2013. And that that was at a time when if you said to a bank or any institution that you worked in crypto, the conversation ended right there. But Jeremy believed and Circle believes that blockchain technology and the internet can actually change the way money moves.

It can change the way money is accessed and transmitted around the world. And that is a vision that we're still building upon 13 years later. So, my job as the chief commercial officer is I sit in these meetings every day with Fortune 500 companies, banks, enterprises, startups. And for the past 5 years, it's followed a very similar script because this idea of what we're trying to do, global money movement, that settles in a second, that settles with at at at a half a penny in cost, is really intriguing. So, the same meetings would happen.

The business team would lean in. The innovation team of these companies would lean in really interested in what Circle is doing. But without a doubt, the compliance officer would would lean forward and say some version of not going to happen. We can't hold crypto. There are no rules.

There is no legal framework. My auditors won't understand it. My board will kill me. Meeting dead. Meeting dead.

And that was the first 5 years of my time at Circle. Meeting after meeting trying to get the world to understand what we were building. But what was really interesting is over the course of the past 18 months, these dominoes began to fall. And the first thing that happened was Circle became a publicly listed company in the New York Stock Exchange. Now, for the first time ever, a stablecoin company is public, it's audited, it's listed for the world to see.

A month after that, in July of 2025, the GENIUS Act passed, creating a framework for for dollar-backed stablecoins, reserves auditable. And so, those same companies that were shooing us out the room and rolling their eyes, now they're they're complete shift in how they thought about things. The question was no longer is this legal? The question is, am I too late? And can you help me launch before my competitors launch?

And so, that's the kind of that that that's the kind of transformation we have seen in just 18 months, and it all starts with trust, and it all starts with an institution that that can really back up what it's what it's talking about in terms of how the money is is attested. So, where are we today in terms of how the internet moves and how blockchain technology can actually help? I will start with a with a with a statement that I think the internet was actually shipped incomplete. Obviously, it has it has helped us move money, move information around the world at great speed, but it shipped incomplete on two factors. Number one, it can't understand any of the information.

And number two, it can't pay for anything. And so, when we think about our our own use cases with the internet, we've been solving for this gap for about 30 years. We have access to information, but that gap has been intermediated by all of you. Think about the last time you went to book travel. You're scanning seven different websites.

You're looking at different prices. You're taking it all into consideration, and then you make a payment. The internet was not built to actually help use judgment on its own, and it certainly didn't have internet-native money. And so, now what's happening today is that these two forces are converging. With AI, AI can actually understand the internet.

It has judgment. It can make actions on your behalf. And with stablecoins and blockchains, now they have internet native money where they can actually use it for the benefit. Now, this is the part where I am actually optimistic is that if these two forces converge upon one another, it can be for the benefit of all of humanity. Because it's real dollars that are being settled on the internet.

So, agents can act on your behalf. But, it's real dollars that people can use on their balance sheet. This is a huge opportunity because agents will not wait for the old payment rails of the of before where money settles in days not seconds or money um costs just as much to send as as the total value itself. There's a lot of opportunity here in terms of how these two worlds are converging. And so, what does that economy look like today?

This is a busy slide, but the one thing that I want you to think about is the 30 cent number. That's the most important number here. The 30 cents is the cost of an average agent payment. Right? And so, this is an agent paying another agent for a data, for for an API call, for inference.

And so, in the past, this kind of transaction couldn't exist because our rails didn't support it. 30 cents. The price of a credit card is typically 30 cents plus two or three percent. That's the price of the entire cargo for agent payments. And so, these kinds of payments are accelerating and in fact, again, 165 million times since spring, agents have been using stable coins, USDC to pay one another.

The number of merchants has accelerated to 10,000 businesses around the world that are allowing businesses to spend money with them. So, this is the bend in the curve that you see. This is an economy that actually hasn't been built yet. This isn't the global economy shrunk down. This is a new type of commerce that that enables agents to to spend money at fractions of a penny with velocity that we've never seen before.

This is what's actually really quite interesting. And so, now there's this new type of customer, an agent. And you have to ask yourself, what matters to the agent? The agent has no brand loyalty. The agent does not care about nostalgia.

What it really looks for are three things. It looks for speed, cost, and instant finality. That's what it looks for. And so far, USDC has been the dollar that agents prefer the most. And when we think about how USDC is constructed, it was actually constructed just for this moment.

It's instant settlement. There's 74 billion dollars of USDC reserves around the world. 99% of all agentic trans- uh uh commerce is happening with USDC today. And so, this is the opportunity where agents now have the ability to make payments with real dollars, with a stablecoin that is auditable, that has a 100% its reserve backed, a public company. And so, this is what we see the world building on top of.

Because the question becomes, once this default is established, it's actually hard to unship it. And we think USDC's really in in a good opportunity to to become the default stablecoin payment for all agentic movement. And so, what I talked about is is really what is happening today. It's a small microcosm. It's small, but again, it's accelerating.

Most of the world don't understand exactly what's happening right beneath our feet. So, I want to talk a little bit about what's next and what Circle believes will happen over the course of the next several years. We think agentic econ- the agentic adoption will actually happen in waves, very similar to how the internet came to be came to be. Today, we are in wave one. And this is agents buying from agents.

These are small micro payments, 30 cents each, happening thousands of times a second for API calls, for inference. This is what we're experiencing today. This is the infrastructure that is being built. It's invisible. It's very similar to when they when the internet launched, how email came to be.

Most people didn't realize it until it became very, very big. That's wave one. Wave two is what's beginning. Is agents buying on behalf of humans. And so there's a scenario where many of you are already using agents to book your travel, where you tell an agent, you give it a budget, and then the agent can go and say, "If you're going to go to Chicago next week and you got to be there at 9:00 a.

m., I'm going to book you the right hotel. I'm going to book you the right plane. I'm going to book you the right restaurant. I'm going to book you the rental car."

You approve that entire transaction. You're not approving seven individual transactions. You give it a budget. And you have human logic around it. So the human isn't in the loop.

The human controls the loop. And we're in the very early stages of agents buying on behalf of humans. The third wave is where the agentic economy takes off and businesses are primarily agents. And you have businesses where maybe there's 10 humans or 100 humans, but tens of thousands of agents working on our behalf, building this global economy. That is when the the bend the curve really starts to bend, because the transactions that happen are unfathomable.

It's hard to predict the number of transactions that will happen in that scenario. But you can you can guess that in that situation, humans will actually not be in the loop as much as they are today. Most of the global transactions will be agent to agent. When all that is happening um on the blockchain, using a publicly uh listed stablecoin like USDC, that's the opportunity. It's hard to fathom what will happen with with transaction volume.

But that's the technology that's being built right today. So the other thing that we really think about at Circle is the payment rails in which these work. And so this isn't a knock on banks. Banks were built for humans. Everything about a bank is about ensuring trust and security.

You show up, there's a name, there's an account, you have to fill it out. All of these things were built for human interaction. And so, these payment rails were actually just not built for agents. And it's not about retrofitting agentic of agentic behavior for banks, it's about creating a new type of payment rail that's really that's really obvious and really interesting right now. And even if you think about ACH, ACH was built in the 1970s.

The time it takes an ACH payment to actually settle, agents have already made thousands of purchases within a second. And so, there's a new opportunity for a new payment rail to be built that goes beyond the traditional infrastructure that was made for humans. Because agents are not going to wait for you to open a bank account, they're not going to wait for banking hours. This is an opportunity for global businesses that are doing commerce around the world where someone in São Paulo is buying your products at 2:00 a.m.

Now you have those funds settle with you at 2:00 a.m. You don't have to wait till this case Tuesday because post Labor Day for those flat money to settle. This is the opportunity. And now for the first time ever, the third belief is we think the foundations are in place.

And there's three things that actually happened, these blockers that came down that set us up for this next next wave of innovation. The first is legal clarity. Once the Genius Act passed, all those companies that I talked about that wouldn't meet with us and said it was impossible, now all of the sudden are thinking about either A, how they launch their own stable coins, you have banks around the world launching their own stable coins, you have card networks like Visa and MasterCard settling in USDC in the hundreds of millions of dollars per month. You have new payment standards like X402 where Stripe and AWS and Google and and and Circle are creating new payment standards for agentic economy. So, this type of innovation is only happening because there is legal clarity.

And that payment standard that I talk about is really going to help unify. So, you know those payments that are being made, they're validated, they're verifiable. That's something the world needs to certainly um catch up to you, and that's that's being built today. So, that foundation that we we've talked about legal clarity and and and open protocol and standards that we have to make sure this next wave is happening is happening today. That's the foundation that's being built.

But, the opportunity is really all in the design. And the way the internet was built, again, it was built with two gaps. It didn't have judgment, and it didn't have internet of internet native money. And so, we have an opportunity in terms of how we design this to fix those flaws. Again, it made sense back in 1975 when the internet was being conceived, you didn't need judgment, and you didn't need internet native money, but now you do.

So, I think the industry has some questions to ask itself. Do we want an open standard where anyone can participate, it's permissionless, and access flows to everyone? Or, do you want a walled garden where, like an app store, there's one company that controls everything, controls who gets to participate. We think that the design philosophies we make right now are going to really dictate what happens over the course of the next several hundred years in terms of how money flows and access to that money. The design quality and the design principles are really what happens, and that is being built right now.

And so, when you think about Circle and our role in this, we play a small part, but from the design philosophy standpoint, there's a platform there's an end-to-end platform, and I won't bore you with every single box. But, for the agentic economy to actually prosper, all three layers have to work together in unison. The bottom layer is the infrastructure layer. Today, USDC is native on over 30 blockchains. We've also launched developer products like the agentic stack, so agents can hold wallets for the very first time.

We're building that infrastructure layer for any developer in the world to build on top of. The second layer is the money layer. This is the regulated money. You need to know that your USDC is always backed 100% and if you need to mint it, you can mint it and if you need to redeem it, you can redeem it. It's the number one regulated stablecoin in the world.

We also have EURC, which is the number one euro stablecoin in the world and we also have our own tokenized money market fund USYC, which is the number one tokenized money market fund in the world. And so, Circle has issued tokenized money for the last 6 years and will continue to do so. The newest layer is the app layer on top of that where builders can build and we've contributed ourselves with the Circle payments network and so this is a network that allows financial institutions around the world to settle in USDC or stable or any stablecoin. And we also have new products that we've launched like Circle FX that allows the conversion from one fiat back stablecoin to another fiat back stablecoin. And so, this is the platform that we believe is necessary from a from a layer perspective starting with infrastructure, then regulated money and then the apps that can build on top of that to really achieve this future that we're talking about where the agentic economy can take over and flourish for all of us.

And so, the vision that we've had is that any business can follow a similar flow. So, let's say you're a logistic business and you wake up and you have uh you have a logistic AI that's helping you and and and you're trying to you're trying to navigate a delay. There's a port delay that you have and so, you're trying to buy actually satellite imagery. Your your AI can wake up at 6:00 in the morning. In fact, it never sleeps so it's never awake.

It's never asleep. Your AI can go to work at 6:00 in the morning and and buy satellite imagery that it finds in an agent marketplace. It could use USDC to buy it 30 cents at a time 4,000 times. And by the time you wake up and you have your first cup of coffee, this is already happened for you. And now a human can actually again enter the loop.

All of it is auditable. All of it All of it is verifiable. And the And the human actually has complete control to ensure that everything that it set out that that that he or she set out is actually being executed by the agent. This is the kind of future that we think USDC and blockchain infrastructure and the agentic economy can bring for all of us. That it's not about the humans not being in the loop, it's about the humans controlling the loop.

And that's the That's the really That's the prize here. And so I want to go back to and end on what I started with, which is really quite exciting. 200 years ago, the Boston businessmen did not know really what they were building. They were building a railroad to move rock for 3 miles starting in Quincy down to the Neponset River. They didn't realize that what they were building would be the bedrock of the American economy where other companies and other builders would build on top of it.

That is really where we are from an agentic economy perspective. We're still in the very very early days in terms of the design that we that we put in place, in terms of the apps that we build, in terms of the roles that we have available, and in terms of who actually participates. That is where we're at. And so the stats that I mentioned for you, the the 165 million transactions, the 80,000 agents that are participating, that is not the agentic economy. Right?

I think the the agentic economy has not yet been built. The stats that I mentioned for you, that's the 3-mile railroad track. We're still in the very very early stages of this, and we're really really proud to be building this with all of you. So thank you very much.

[applause]

If anyone has any questions, I have a mic back here for you.

Sure. All right. Be easy.

Awesome. Thanks a lot, Roman Beck Bentley University. Fantastic presentation.

Thank you.

Um we are in Quincy, but I'm asking about the Q-day anyway. Um as it fits quite well, kind of agentic payments and building the rails for the next century or hundreds of years. You mentioned design principles and everything. Super important. How ready is Circle for uh the post-quantum era?

Um I don't know is the right answer. Uh you know, I think the investments that we are making in the post-quantum era in terms of the agenting workforce to give you a sense of scale, we haven't stopped hiring or or or or you know, decreased the number of people. We've actually accelerated. And so, the work that we're doing for inference for for example, and the work that we're doing on the on the agentic side is is is incredibly uh aggressive. And so, again, it goes back to really what I was saying.

The design principles that you put in place today can actually become a guardrail and a framework in terms of how decisions are made. We do not know what the world will look like when transaction volume accelerates by a million X. What does compliance look like in that situation? How do you ensure that money movement from money movement from border to border is actually safe and secure? What kind of standards do you need to have in place?

And I think the post-quantum era represents that opportunity to accelerate by again, a million X. Um the framework that we put in place today, I think will dictate how ready we are, but that is a huge level of involvement investment for us. I I don't assume that the design principles will will will be sta- will we etched in ink forever. I think the design principles are a good set of principles to start with, and then you you evolve over time. Um and so, you know, again, I'm not a crypto I'm not a AI researcher.

I'm not a cryptographer. I I um I don't necessarily uh from from my perspective and circles, we're trying to get ready for that eventuality. Um and you know, we're building with with everyone here.

Hi, my name is Hamid Pishtadian. I'm the CEO of uh SQE. We actually answered exactly the question that you're asking. Uh and I just wanted to say thank you for the talk. I agreed with everything you said.

Uh almost everything in terms of security that you're asking, uh it's exactly what we implemented to make a million transaction per second a reality and we are ready. And we can just help anybody to do that today.

Thank you.

Hi, thank you for the talk. Um I want to pose a thought question because I think the internet was built without another layer. So, I appreciate you saying it was shipped incomplete. Uh and that's the authority layer, uh which might sit somewhere in your stack because when you think about agents who are doing things on your behalf, we run into exactly the principal agent problem. Um and so, I wonder where you think there will be the layer that says, yes, I did in fact want that transaction and I will want to pay for that.

I think that's the authentication layer that that I talked about with in in terms of A, you know, authorizing who you are, making sure there's a universal way of of people being able to verify themselves and and and and and prove they are who they are. And B, uh again, the the the standards that are being built today, whether it's the X402 Google standard, Stripe's new authentic checkout standard, all of these are being built to solve that exact problem. So, if I give a budget to my agent and my agent is acting on behalf of me, and then I go to sleep and I wake up in the morning, all of those people, all those businesses that participated in that flow, need to know that that agent's actually acting on on my behalf. That's a incredibly big unsolved problem that hasn't been solved, but I think, you know, what gives me hope is that there are other businesses around the world that are working on that problem today.

So, thank you. I I'm also working on that, and my contribution for you to think about is that authenticating probably not enough, uh because we have authentication for people, but once a session is authenticated, once a person is authenticated, there's still opportunities for other people to imitate us. So, that piece, that cybersecurity piece of knocking down to the task action, is something I'd love to talk to other people about, rather than just identity.

Thank you.

Uh thank you. A very exciting talk. Um it's nice to look forward and see at least your vision of the future. Um I have a question, and maybe it doesn't have an answer, but you talked about minting and redeeming, and you talk about cross-chain transfers. Does um does uh you USDC take on the status of a currency when it when it represents the transfer of assets?

It's a currency today. It takes it is a payment stablecoin, so it's only as good as the trust behind it. And so, it's a it is a digital dollar uh that lives on the blockchain. And so, um you know, in in in some ways it is a is a it is a digital currency uh backed by the US dollar.

Right. So, excellent. There's obviously technology layer considerations. There's operational level considerations and then there's governance and you're you're suggesting that the governance piece is being worked on right now. That's what you you kind of mentioned when you talked about the genius act.

Is that right?

The genius act allows a framework for for companies to actually get involved in stable coins. A lot of the governance in terms of how a dollar moves from one chain to another chain is auditable and it comes down to the to the technology that's being built. So, Circle has spent an enormous amount of resources in terms of building technology that allows money to move from one chain to another chain seamlessly. So, that's a huge part of our investment.

Thank you. I appreciate it.

Thank you.

We have time for one more.

Thank you. Question. Automated agentic investing into structured products is a fascinating application of what you're talking about today, but I'm curious about what Circle is thinking about the trust problem and then the evaluation of the quality of those structured products. Thank you.

Yeah, the trust problem is a I think it spans many different vectors. It's not just the agentic payment. It it has to do with everything and and we know we've spent 15 years being a boring company so companies and and individuals can trust us. You know, it has to do with everything we've done with being regulated. We've walked through the front door of every regulator around the world to make sure that regulators understand what it is we're doing and why we're doing it and we've done that at the sacrifice of growth.

We're the most licensed stablecoin in the world and I think first it starts with working with local governments to ensure that they are aware of what's happening so their citizens and the businesses within their jurisdiction can actually feel safe. So, first is there's a there's a verifiable trust and social contract between humans. And then the next thing I think it it goes to the technology that's being built and ensuring that that technology is something that can evolve over time like the gentleman was saying here, but ensures that again the people that are controlling that technology are the ones that have that track record. And so that is something that we tried to build over the course of the last 15 years. We've tried to become a company or be a company that the world can trust because ultimately this is real money and money can do a lot of things as as we obviously know from from building a new economy to even being an expression of love.

And so that is something that we care about a lot at Circle. We spent a lot of resources doing that and hopefully it's something that we can continue on.

Thank you.