Why Your AI Agent Needs a Wallet: USDC and Nanopayments — Harshal Bhangale, Circle

AI Engineer · 20 min · 179 sentences · from YouTube's caption track

Each timecode opens YouTube at the start of that sentence. Line anchors (#s42) are the cue ids in the WebVTT, and every line carries its start and end seconds. All transcripts has every talk, and the whole corpus as one file.

00:01[music]
00:12Good morning, everyone.
00:15So, um I've had the pleasure of uh meeting a lot of folks at uh the booth here over the last couple of days, and the interactions have been great, uh but
00:27one question kept uh coming up, which was Circle is a stablecoin company.
00:32So, why is a stablecoin company at an AI engineering conference?
00:37And well, uh the answer to that is fairly simple, because Circle issues uh USDC, which is the world's largest regulated stablecoin.
00:49And we've built over the years expertise in making payments simpler and cheaper.
00:55And it turns out that's one of the bottlenecks for your AI agents.
01:00So, when you often think about making AI agents smarter, we think about better models, uh more tool calls, complex orchestration, but where your agent actually in practice actually halts is when it hits a paywall
01:19or when it has to pay for something.
01:22Then you have to step in, either create an account, sign up, or, you know, manage API keys.
01:29So, that's where uh the limitation is, and that's the gap that I want to talk about today.
01:37So, yeah, I'm Harshal, and I'm an engineer on the agentic product team at Circle.
01:45So, let's sort of take a step back and see how we got here.
01:49We believe that the agentic economy is already here.
01:52So, how are uh you know, uh we In 2023 we interacted with agents via like prompts, uh your ChatGPT and stuff like that.
02:03We got a little bit better.
02:05In 2024 we built workflows.
02:072025 was all about MCPs, skills, and orchestration.
02:12And we believe this year, 2026, is when agents actually start paying for services that they want.
02:19And the signals are promising.
02:22Just in the last 30 days, uh agents have transacted with paid API endpoints, uh and the volume is about like $24 million.
02:32Uh over X102.
02:35And 99% of it has been settled in USDC.
02:40While this number may look small in terms of volume in the broader landscape, it's only about to get larger.
02:48So, what is X102?
02:50Um essentially, X102 is a way where agents can uh pay for their the resources that they want because the server essentially returns a 402 header and with the details
03:04of how they want the payment to proceed.
03:07And then the agent just signs um an authorization from the crypto wallet and then pays for the resource and retries the request again.
03:15So, that's in X102 in a nutshell and we'll see that in action in a live demo.
03:21But then, the question is why do the traditional payment rails not work and why do the agents get stuck?
03:30So, the answer is simple, uh because for the last 30 years we built the internet around one customer and that was humans.
03:39Right?
03:40Um so, we built payment schemes, uh monetization strategies, all catered towards how humans interacted.
03:47So, you had your sign-up flows, add your credit card, uh put in your information, and manage your API keys, but agents just don't function that way.
03:58Agents want to like come in and grab that piece of data, resource, compute, inference, whatever, and just uh you know, and they have the ability to even like consume knowledge at a scale that humans just cannot.
04:12I'm sure like you have seen sessions where it's able to like just scrape through hundreds of web pages, and um then it just like stalls because it can't like reach a particular endpoint,
04:23and it skips over it.
04:25So, then you could think, "Oh, yeah, just give it a credit card."
04:30or something like that.
04:31So, while that is possible, um these agents, uh because they consume so much data as they go, it's um they pay in fractional amounts, so tiny amounts, uh but at a very high frequency.
04:47Uh and credit cards and their fees are just like not sustainable for this kind of like economic model.
04:53Um you cannot pay like 3% uh each time an agent tries to make a one-cent transaction.
04:59And the reason these transaction amounts are so small is because on the sell side or on the merchant side, they've realized that these paywalls were actually catered for humans,
05:09and now there's an entirely different customer base which is trying to like access their data, and they just want like a subset of the data.
05:16So, you could monetize that as a seller by just offering that, by wrapping that in a uh you know, in a paywall, and saying, "Hey, I take one, pay me 10 cents and grab this data."
05:29And that's why these are tiny um microtransactions, but highly frequent.
05:35So, because the agents are just like making these API calls constantly.
05:41So, what do these agents need?
05:44The agents need um payments to work like the internet.
05:48So, they have to be real time, low cost, programmable, and always on.
05:57And that's why we've built the Circle agent stack.
05:59So, it's the full stack platform for the agent tech economy.
06:02And what we mean by that, let's look at it with a live demo.
06:07Um yeah, wish me luck.
06:12Okay.
06:14So, I'll just explain uh what I'm trying to if I can get the terminal.
06:24Okay.
06:26Okay.
06:29Let's Mhm.
06:34Pull it.
06:36Yeah, I don't know.
06:40Excuse me, can I Yeah, I tried but it's just not I don't know where the placement of the desktop is.
06:52Can you escape out of this?
06:53[snorts]
07:00Okay.
07:06Thank you.
07:07So, so what we have over here are two sessions.
07:11One is your regular cloud code and on the other on the right hand side is a cloud code that comes equipped with Circle agent wallet.
07:18So, it has a wallet which is funded and has the ability to like make uh has the ability to pay for premium content.
07:25And now uh let me actually just uh quickly uh add a task to this.
07:36Yep.
07:38And run it.
07:40So, I'll explain what I'm trying to do over here, which is Yeah, so So, the task I'm giving Claude on and it's the same task on both the terminals is
07:51plan my trip for the FIFA World Cup final.
07:55So, just, you know, give me summary of flights, hotels, logistics.
08:02Also, like what are the odds of like my favorite team Argentina being in the final?
08:06Who is it going to play?
08:07And stuff like that.
08:09Um and also what's the ticket prices in the secondary market?
08:13How is the experience of other people who've actually been to the stadiums?
08:18And are there any FYIs and stuff like that?
08:22Grab all of that.
08:23Send me an email.
08:25And if possible uh make me a make a phone call and tell confirm that all of this has been researched and you know, sort of brief me on the summary.
08:34So, we'll see like what's happening on on both the terminals here.
08:38And Sorry, it's a bit hard to navigate.
08:42So, as you can see, um the left the vanilla Claude code has spun out like a bunch of sub-agents.
08:49And it's going about doing its research.
08:53Similarly, on the right-hand side um it's actually like uh going through the wallets and making like um paying for all of these uh premium content and we'll see like what the results of it are.
09:12So, as you can see, it's trying to make a phone call to find sorry, an API call to Stable and Rich.
09:20Um and it's paying from its wallet.
09:22It's set up a guardrail of like, "Hey, max amount is 15 cents."
09:27The beauty of uh having an agent with a wallet is you sort of build these guardrails into the wallet.
09:37And you don't have to as a human approve every single transaction because that would just not scale because these agents are just making these ones and five cents, 10 cents transactions.
09:46You want to like enforce those guardrails which is hey, this is the max amount you can send spend per session or this is the max cap you can do per day and stuff like that.
09:55So, the agent is still spending within the guardrails that you set but it is autonomous enough to like make these individual API calls.
10:07So, let me just look at what the vanilla one is looking up and doing some research for hotels.
10:22Yeah, um flights are coming in.
10:28Also like in the prompt like try to like tell the agent to wrap this whole thing up within 6 to 8 minutes in the interest of time.
10:35So, let's see if we're able to like get an email and a phone call done in the same time.
10:51Let me also walk you through what the agent's doing if I can grab.
11:05Yep, as you can see over here the agent was able to like make an API call to the poly market data via a provider called Block Run and it provided a query
11:20and again set the max amount guardrails and is now passing all the results.
11:53All right.
11:54So, let's go back to the main subject main agent here.
12:23Yep.
12:23And this is where I think you'll see on the left-hand side that the agent gets stuck because the agent just does not have the ability uh it cannot send out an email
12:34natively.
12:34So, what it's doing is it's using my uh it's just like adding a draft into my Gmail account uh that is logged in, but it just cannot send it to someone.
12:44Uh whereas the other agent on the right-hand side will just be able to use um a provider, pay it, and send an email.
13:00Now, let's see like this one's uh the the other agent is also like uh is about to like send an email and place the call, so let's see.
13:11Does.
13:30Okay.
13:32So, as the I think the left agent, the vanilla cloud goods finished, and as you can see, it's sort of like made a confession that yeah, it cannot have It does not have the ability to make a phone call.
13:46So, it's just like providing me the highlights over here in the terminal itself.
13:50Um and now, uh whereas on the right-hand side, the email's been sent.
13:57Now, let me try to actually show you the email that we got, I wish.
14:10Grab this somehow.
14:19Okay, I'm going to make make the phone call.
14:25Okay.
14:35Just going to grab my thing here, and then let's look at the email.
14:46So, yep, there we go.
14:52So, I got this email.
14:54Oh, this is the old one.
14:55Let's see the new one.
14:59Yep.
15:01This is the one that we just received like 2 minutes ago.
15:05And yeah, it has all the details that it was able to like find, including like getting to the stadium, open stadium and maps.
15:15It was able to like, you know, find all of these.
15:20And then, what to expect by looking at like Reddit tickets, etc. So, I'll wait for the call but in the interest of time it takes sometimes a few minutes for the call to come in
15:37but let's see and I'll move on.
15:40So, how does this work uh in the background?
15:45Is Yeah, let's get back to this.
15:50Let's Yeah, so so as you can see like the agent on the right hand side with the wallet was able to just make a bunch of these API calls and pay for it.
16:07While blockchains make all of these things theoretically possible, there are there are also some bottlenecks because uh even the most efficient blockchains have a gas fee and these small
16:19transactions just don't scale because the gas fee will again be a significant sort of fraction of the actual transaction.
16:28But it is important to have gas fees because they prevent the network from spam and abuse.
16:33Um and blockchains also have some throughput limitations because there are other use cases that run on blockchains and block space is shared infrastructure.
16:42So, the problem for agents is unpredictable latency and degraded performance under load.
16:47So, what we did for that is essentially uh we built a new um infrastructure layer on top of our intra product called gateway.
16:59So, it's called nano payments.
17:00It's built for sub cent transaction sizes for as low as one micro cent.
17:05Uh it supports um it is gas free for the seller and it's instantly cross-chain.
17:12And the way it works is you essentially you just fund um your wallet and then Oh, sorry.
17:20While I explain this, let's listen to the actual call that came in.
17:23And
17:24Hi Harsha, here is your World Cup final trip
17:27I don't know if it's part of it.
17:28at MetLife Stadium.
17:30Fly SFO to either EWR or JFK non-stop.
17:34SFO is showing live arrival delays today, so build buffer.
17:37Awesome.
17:38Can you tell me how do I get to the stadium from my hotel?
17:45On match day, take NJ Transit to Secaucus Junction, then the Meadowlands Rail Spur direct to the stadium.
17:51Budget about an hour before receipt.
17:53Thank you.
17:54I'll read the rest of it in my email.
17:56Thank you.
18:01You're welcome.
18:01Have
18:03All right.
18:03So, that was just a quick demo of like how you can have like AI agents.
18:09Um if they're equipped with a wallet, they can do a lot of things with the as opposed to an agent that does not have a wallet.
18:16So, I have a couple of more minutes, so I'll just walk you through how this technology works underneath.
18:23So, if you have a wallet and you funded it with USDC, it's very easy because Circle works with a bunch of providers to on-ramp your actual US dollars into USDC.
18:35And from there, you can deposit the funds into a smart contract.
18:38The next thing it does is the agent just has to like sign these off-chain authorizations, which are essentially cryptographic signatures saying, "I am paying this address this particular amount of money."
18:53And then the server just has to like relay this to Circle and within a few hundred milliseconds, this the server knows the merchant knows that, you know, the user has the funds
19:06and is able to release the actual resource that the agent requested.
19:10And with this, you avoid the issue of like you know latencies and stuff that are associated with actually settling every single transaction on chain and this the agents are able to like pay
19:23for things at the speed at which they operate.
19:26So in conclusion the the the way the stack would work is you have Circle agent wallets which give you the ability to equip your agents with wallets.
19:38The agents have can now hold their money, spend that money autonomously but within the guardrails that you set.
19:45The wallets enforce those guardrails and then on the sell side the merchants are able to like just wrap their endpoints and resources and monetize it with a few lines of code using our SDKs.
19:59Then USDC and nano payments is the layer underneath it which sort of helps settle these transactions at the speed at which agents operate which is sub second.
20:10So and that this is how it scales and you can give it a try yourself by going to agents.circle.com.
20:20It's just a couple of clicks and you will have your agent equipped with a wallet and ready to like you know make these phone calls and things like that.
20:31So yeah, thank you.