AI Devs Skeptical of Crypto, But Stablecoins Key to Agentic Finance

Stablecoins agentic finance is shaping up as one of crypto's biggest real-world bets — and the people building it know they have a perception problem to solve first. Crypto insiders are convinced that dollar-pegged tokens on public blockchains are the only logical payment rail for a world where AI agents transact autonomously, around the clock, across borders. The AI developer community, though, largely wants nothing to do with them.

What Is Agentic Finance and Why Do Stablecoins Fit?

Why are stablecoins better than credit cards for AI agents?

Agentic finance refers to a category of commerce where autonomous AI agents — not humans — initiate, process, and settle transactions. Credit cards and traditional bank wires weren't designed for machines running 24/7 and firing off thousands of micro-payments in fractions of a cent. Stablecoins, by contrast, were. The programmability and composability of stablecoins agentic finance infrastructure mean that coins can be set to transfer only when specific conditions are met — and multiple actions can be chained together automatically on receipt of a token.

Dante Disparte, Circle's chief strategy officer and head of global policy, put it plainly in an interview. Programmability and composability are the two essential features that make stablecoins suited to agentic commerce, he said. The blockchain ledger itself becomes the shared reference point that agents consult — a neutral, always-on record that no single company controls.

Firstly, you have to be able to exploit the otherwise really innocuous features of stablecoins, which is programmability and composability. Number two, where the stablecoin lives, the physical blockchain ledgers themselves, are the common reference point the agents will turn to.

The AI Developer Community Has a Crypto Problem

Here's the uncomfortable part nobody in the stablecoin camp wants to lead with: a lot of AI builders genuinely dislike crypto. Peter Steinberger, creator of the AI agent OpenClaw, is publicly opposed to the space — so opposed that he declined to comment for this story at all. That kind of hard pass isn't an outlier.

Sean Neville, co-founder of Catena Labs and also a co-founder of Circle, has worked on both sides of this divide. He's unusually candid about what he's seen. Speaking in an interview, Neville acknowledged that engineers in the AI community carry real skepticism — driven largely by the industry's history of memecoins, outright scams, and projects that burned retail investors. The challenge, then, isn't technical. It's reputational.

"I think stablecoins have achieved some escape velocity," Neville said, "but the AI developer community in particular has a negative view of crypto, because of things like memecoins and Ponzi schemes and whatnot." That's not a rounding error in perception — it's a structural headwind that the agentic finance sector has to work through before the infrastructure even matters.

Catena Labs, Coinbase, and the Race to Set the Standard

Catena Labs, co-founded by Neville, raised $18 million in seed funding last year, with a16z leading the round. The company is focused on agentic finance infrastructure — specifically, how to give AI agents a financial identity, programmable spending controls, and enforceable policy guardrails that prevent rogue bots from causing chaos.

The harder problem Catena is navigating: regulated money transmission in a landscape of countless bots with no financial identity. The goal is to keep bad actors out while letting authorized agents operate within strict parameters. "The way to handle that is programmable money," Neville said in an interview, "because we can leverage cryptography to ensure verifiability and auditability and so on."

On the protocol side, Coinbase has led engineering on x402 payments protocol, which embeds stablecoin payments directly into HTTP interactions — essentially making payment a native part of how agents browse and transact on the web. Erik Reppel, head of engineering for Coinbase Developer Platform and an x402 founder, isn't losing sleep over month-to-month transaction volume comparisons. He's focused on something bigger: the collapse of the internet's advertising model.

I think the thing people haven't quite realized is that we're going to break the fundamental economic model of the internet, moving from browsers and you visiting the website of the person who's publishing content, to consuming things through your agents and your chat interface.

Nano-Payments, Protocol Fragmentation, and the SSL Problem

The transactions being discussed here are tiny. We're talking nano-payments — fractions of a cent — fired off at high frequency between agents, with a human somewhere in the background. Visa and Mastercard weren't built for this. Spinning up virtual cards is technically possible if a developer has a relationship with a card network, Reppel acknowledged, but it's clunky.

"Anyone can program stablecoins," Reppel said. "Anyone in the world can spin up as many wallets as they want, and then just use wallets as the way to fully isolate funds for an agent. What we want is agents to have isolated, programmable funds, where your agent can't spend into your credit card limit and can't access your credit card." That isolation is the killer feature — not just the payment speed.

The more urgent structural risk isn't whether stablecoins can handle agentic commerce. They can. The risk is fragmentation. Multiple protocols are fighting for position, and if AI agents can't agree on how payments work, marketplaces stall before they launch. Neville wants to see something like an SSL equivalent emerge for agents — a standard nobody owns, that every builder can adopt. SSL, or Secure Sockets Layer, is the technology that encrypts connections between web servers and browsers. That kind of universal, open standard would let the whole sector move together rather than fragment into incompatible silos.

Does the Stablecoin Bull Case Actually Hold Up?

The argument from Disparte, Neville, and Reppel is internally consistent. Stablecoins are programmable. They run 24/7. They settle cross-border transactions faster and cheaper than traditional banking infrastructure. In a world where AI agents are operating continuously across jurisdictions, that matters — a lot. Add regulatory clarity finally arriving in the U.S. for stablecoins, and the macro setup looks favorable.

But here's the thing the crypto camp glosses over: the AI developers building these agents don't have to use stablecoins. They can route through card networks, use bank APIs, or wait for whatever payment layer emerges from the major AI platforms. The stablecoin pitch only wins if the tooling, the compliance layer, and the developer experience get good enough that the reputation baggage stops mattering. That's not guaranteed.

Circle's CRCL and Coinbase's COIN are obvious winners if agentic finance scales on stablecoin rails. Whether the AI development community meets them halfway is the real open question — and nobody on the crypto side gets to answer it for them.