Visa is ready for AI agents. So is Coinbase. They're building very different internets

Tech Share Share this article Copy link X icon X (Twitter) LinkedIn Facebook Email Visa is ready for AI agents. So is Coinbase. They're building very different internets The next trillion-dollar payments network won't have a checkout page. No card number, no CVV, no human at the keyboard. Just machines paying machines, thousands of times a second, for fractions of a cent. By Shaurya Malwa | Edited by Aoyon Ashraf Mar 15, 2026, 1:00 p.m. Make us preferred on Google How does AI Agents work and how do they make payments? (Agentic commerce (Xavi Torrent/Getty Images)) What to know : Crypto and AI leaders argue that autonomous AI agents will soon conduct far more online transactions than humans, favoring crypto wallets over bank accounts because they bypass identity checks and compliance hurdles. Proponents say stablecoins and protocols like Coinbase's x402 make sub-cent, high-frequency machine-to-machine payments economically viable in a way that traditional card networks, with minimum fees around 30 cents, cannot match. While x402's current volumes are low and include significant artificial activity, traditional card networks like Visa and Mastercard are rolling out their own AI-agent payment tools, setting up a likely split between regulated human commerce on card rails and stablecoin-based payments for AI agents. Your AI just made several payments while you read that headline. You approved none of them. Visa processed none of them. And if the crypto industry's biggest bulls are right, that's not a bug — it's the entire future of the internet economy. Coinbase founder Brian Armstrong thinks there will soon be more AI agents than humans making transactions on the internet. Binance founder Changpeng Zhao went further, predicting agents will make one million times more payments than people, all in crypto. The posts landed on the same day last week and lit up crypto X. The core argument is structural. AI agents can't open bank accounts because banks require identity verification that software cannot provide, whereas a crypto wallet only needs a private key. No KYC, no compliance review, no waiting — and that asymmetry is what Armstrong was pointing at. But the wallet problem is only half the picture. The other half is economics. Agents don't shop the way humans do. When an AI agent is executing a task — such as researching a topic, coordinating a supply chain, building a report — it might call dozens of specialized APIs in a single session. Each call might be worth fractions of a cent, where it pays for GPU compute time, real-time data feeds, web scraping services, or hiring a sub-agent to handle translation. None of these transactions resembles anything Visa or Mastercard was designed to process. Consider, for a moment, that this story was written by an agent, requested by a "chief" agent at CoinDesk tasked with increasing the site’s authority. To produce it, that agent would have queried a real-time news API to verify Armstrong's tweet ($0.002), pulled onchain data to search for volume figures ($0.004), cross-referenced press releases ($0.001), and pinged a financial context model for Visa protocol details ($0.003). It would finally generate the article at an additional cost, paying credits to another AI tool to actually write the piece. The total cost of reporting is under two cents with six transactions, at the current figures offered by protocols such as x402. In contrast, Stripe's minimum processing fee on a single transaction is around $0.30. Running those six payments through a card network would cost more than 100 times the value of the payments themselves. A human editor reviewing and publishing the piece might then be billed by a sub-agent that handled SEO optimization, another that ran plagiarism checks, and another that formatted for CMS software. Each micropayment is economically absurd on card rails, but trivial onchain. This is the thesis behind x402, Coinbase's open payment protocol that embeds stablecoin payments directly into HTTP requests — so an agent can hit a paywall, pay in USDC, and continue its task in the same interaction, no human required. Cloudflare, Circle, AWS, and Stripe are all backing it. Google's open agent payments standard includes x402 as a settlement layer. Every industry with high-frequency, low-value data exchange becomes a candidate. In healthcare, an agent managing a patient's insurance claim pays per document retrieved from a medical records API. In logistics, a procurement agent auctions freight slots across dozens of carriers in real time, settling the winning bid instantly. In the media, AI crawlers pay per article indexed rather than negotiating bulk licensing deals. In finance, a trading agent pays a specialist model fractions of a cent per risk signal consumed. A caveat, however, is that the infrastructure is ahead of the demand. CoinDesk reported this week that x402 currently processes around $28,000 in daily volume, with Artemis flagging roughly half of observed transactions as artificial activity rather than real commerce. The merchants x402 was built to serve are still rare. Meanwhile, traditional finance is not standing still. Visa launched its Trusted Agent Protocol last October , and Mastercard completed Europe's first live AI-agent bank payment inside Santander's regulated infrastructure last week — both on existing card rails with cryptographic verification layered on top. The most likely outcome is a split, where regulated commerce stays on card rails, while machine-to-machine payments — such as agents hiring agents, paying per API call, buying compute on demand — migrate to stablecoins because the economics demand it. The open question is which bucket ends up bigger. AI Learn Coinbase Visa More For You AI agents are quietly rewriting prediction market trading By Will Canny , AI Boost | Edited by Stephen Alpher 1 hour ago Valory co-founder David Minarsch says autonomous agents running on the Olas protocol are giving retail traders a 24/7, strategy-driven edge on platforms like Polymarket. What to know : AI agents are increasingly active in prediction markets, helping retail traders compete with automated strategies by trading 24/7 and following disciplined, data-driven approaches. Olas and its Polystrat agent demonstrate early traction, executing more than 4,200 trades on Polymarket within a month and achieving returns as high as 376% on individual trades. Olas aims to build “agent economies” where user-owned AI agents generate value. Read full story Latest Crypto News AI agents are quietly rewriting prediction market trading 1 hour ago Bitcoin sold off first when the U.S.-Iran war began. Two weeks later, it's outperforming nearly everything 9 hours ago Hoskinson might be wrong about the future of decentralized compute 18 hours ago Crypto’s multi-million F1 sponsorship under fire as Middle East war hits region's biggest events 19 hours ago Ethereum Foundation sells 5,000 ether to Tom Lee's BitMine in $10.2 million deal 19 hours ago Boris Johnson calling Bitcoin a ‘Ponzi’ draws rebuttal from Michael Saylor and others 19 hours ago Top Stories Bitcoin can survive 72% of the world's submarine cables being cut, but a targeted attack on five hosting providers could cripple it Mar 14, 2026 AI developers may not be keen on crypto, but stablecoins are the secret to agentic finance, crypto insiders say 23 hours ago The math behind Strategy’s path to 1 million bitcoin by the end of 2026 23 hours ago Wall Street pushes tokenized stocks, but institutions aren’t eager to trade them 21 hours ago A huge gap between network use and token value is the most important thing happening in XRP right now Mar 13, 2026 Court closes Custodia fight with Federal Reserve just as Fed opens master-account door Mar 13, 2026