Translated by: Block unicorn

Preface

Visa's entire business model is based on betting on human behavior. It concerns human consumption and psychology. The reward points you accumulate, the fraud protection you rely on, the coveted Centurion Card you dream of, and the zero liability policy that reassures you when using ATMs abroad all exist not because of difficulties in fund transfer, but because of human anxiety, the pursuit of status, and a lack of attention to terms and conditions. Visa has capitalized on this cognitive disparity to create a $500 billion company.

However, AI agents do not possess these traits.

They do not accumulate points, do not seek fraud protection, and do not crave black cards. They have only one directive: to complete tasks. And when the task involves payments, the agent performs complex calculations that humans would never bother to make: the cheapest path, the fastest settlement, the lowest fees. Every time it does so, automatically, devoid of any emotion.

Last month, an article titled "The Global Intelligence Crisis of 2028" on SubStack caused Visa's stock to plummet 4% in a single trading day, MasterCard dropped 6%, and American Express fell 12%. The report was labeled as a "scenario analysis" rather than a "forecast" (as stated in the original). However, the market did not buy it. Technological conclusions are irrelevant. The issue is that by 2027, agents will bypass trading centers and use stablecoins for settlement. Visa spent fifty years perfecting its products, yet now its customer base is being replaced.

In machine-to-machine commerce, a 2-3% exchange fee is clearly a target. This conclusion from Citrini Research is its core argument. This does not mean that artificial intelligence will destroy Visa tomorrow. Rather, the fee structure that Visa has relied on to build its commercial empire is essentially a tax on human irrational behavior, while traders themselves are entirely rational. This is the essence of Visa's existence.

What is Visa Selling?

To understand why this is important, you must know what the exchange fees are actually used for.

When you shop with a credit card, the merchant pays a 2-3% fee to the credit card network and your issuing bank. This fee covers your reward points, fraud protection, shopping insurance, and dispute resolution services. The entire consumer value proposition of credit cards is borne by merchants, and merchants ultimately pass these costs on to consumers by slightly raising the prices of goods. It is a well-established and stable system that has run for fifty years because consumers willingly absorb all these costs, albeit they do not pay directly.

AI agents do not need these. They do not dispute fees nor request refunds. The justification for this fee lies in its ability to prevent human errors, fraud, and impulsivity. If there is no human involvement in a transaction, this fee becomes entirely meaningless.

American Express embodies this issue most typifyingly. Its customers are high-income, high-spending, and aspirational premium cardholders. Its annual fees are higher than Visa or MasterCard precisely because its customers are willing to pay for status and privileges. This model presupposes that purchasing behavior is human-driven, and customers choose American Express over Visa because the access to lounges is worth the expense. However, agents do not voluntarily choose American Express; they only look for the cheapest solutions to complete transactions. In a world where software controls credit cards, high-tier membership levels do not exist.

A business routing model led by agents that bypasses exchange fees poses a greater risk to credit card banks and single-issue issuing institutions that heavily rely on 2-3% fee income and build entire business segments around merchant-subsidized reward programs. Visa and MasterCard possess adaptable network operations. However, issuers that have built their entire profit and loss model around exchange fees and reward programs have no way out.

The Week Where Everyone Ships at the Same Time

The Citrini report and the launch of infrastructure projects coincidentally released within the same three-week span.

Tempo officially launched its mainnet last Wednesday. The payment blockchain jointly developed by Stripe and Paradigm is designed for high-volume stablecoin settlements, launched concurrently with the Machine Payment Protocol (MPP). MPP is an open standard that allows AI agents to autonomously pay service fees without manual approval for each transaction. The protocol introduces a session mechanism. Agents only need to authorize a spending cap once to continuously make micropayments while consuming services such as data, computation, or API calls. Payments utilize OAuth authentication. Users authorize budgets, and agents can spend. The entire process does not require using a card at each step.

Anthropic, DoorDash, MasterCard, Nubank, OpenAI, Ramp, Revolut, Shopify, Standard Chartered Bank, and Visa are all listed as design partners for Tempo. The entire payment and e-commerce ecosystem recognizes this structural change.

On the same day Tempo launched, Visa's cryptocurrency division unveiled a command-line interface tool for AI agents to make payments through terminals without API keys, without accounts, and without manual authorization. Visa refers to it as "command-line commerce"—machines can transact without human intervention.

MasterCard agreed to acquire stablecoin infrastructure startup BVNK for $1.8 billion. Circle launched Nanopayments on the testnet, which enables transactions below a cent and gas-free USDC payments for agents requiring no accounts or credentials to use pay-per-use APIs. Sam Altman's World project launched AgentKit, allowing agents to carry crypto proofs to demonstrate they represent a real person, with the toolkit integrated directly into Coinbase's payment system, enabling the platform to validate agent identities without obstructing legitimate transactions.

In my view, what happened last week was that companies are competing to become the new Visa to prevent Visa from realizing what it has already lost.

The Obvious Paradox

There is no disputing that Visa has not stalled.

It has participated in the development of Tempo’s Machine Payment Protocol (MAPPS), launched the Visa Crypto Lab, and its head of cryptocurrency wrote in Fortune magazine explaining how agents could use credit cards to pay via new standards. MasterCard is investing $1.8 billion in stablecoin infrastructure. Stripe has acquired Bridge and Privy. Existing enterprises have recognized this shift and have been preparing long before the new infrastructure arrives.

Visa's argument is that it can expand its track into agent-driven commerce before establishing itself as irrelevant in such a realm.

This claim is not entirely wrong. Stripe processed a total payment volume of $1.9 trillion in 2025, a 34% year-over-year growth. These companies are not shrinking. The distribution advantage of card organizations' networks is hard to replicate. I admit I am reluctant to say this publicly because historically, whenever someone makes such an argument, a new product is released that makes them look foolish.

So, the flaw in this argument is here: Visa's distribution advantage is built on relationships with merchants and consumer trust. Merchants accept Visa because consumers hold Visa; consumers hold Visa because merchants accept Visa. The entire cycle relies on humans. Once agents become the primary buyers in a significant business domain, this flywheel will slow down. Agents have neither brand loyalty nor wallets. All they possess are budgets and directives. The cheapest and fastest line wins their business, and there are zero switching costs.

I want to accurately convey where we stand, as the current public sentiment development is outpacing the data itself.

Although the ecosystem surrounding x402 is valued at around $7 billion, on-chain data indicates that the daily transaction volume on the protocol was only about $28,000 last week, with most of it coming from testing rather than actual transactions. This number is astronomically different compared to Visa’s daily transaction volume.

The transaction volume for x402 has surpassed 50 million. Although individual transaction amounts are small, the number of transactions indicates that the infrastructure is being utilized. Developers are building upon it. The merchant-side services that accept agent payments are also expanding. This is how payment networks start.

McKinsey estimates that by 2030, AI agents could facilitate $3 trillion to $5 trillion in global consumer transactions. This estimate may be correct or overly optimistic. But there is no dispute that the agent-driven business model has not yet achieved widespread adoption. Merchants building native agent services, companies that treat agents as primary buyers, and transaction volumes that can genuinely test the economic viability of transactions are still developing.

The reason Citrini's report caused market panic is that it simulated a series of credible events. MasterCard's Q1 2027 earnings report will not attribute a slowdown in transaction volume to "agent-driven price optimization." At least not yet.

The first to be affected are micropayments for AI infrastructure, not consumer commerce.

Agents completing research tasks invoke specialized data APIs hundreds of times per session. Each call costs only a fraction of a cent. Over a week, these calls could generate $40 in revenue for the developers operating the service. Credit card networks cannot respond to this situation. The economic model for minimum transaction amounts does not work. The merchant onboarding process does not work. The fee structure does not work. Such business models cannot operate within Visa's framework. They require a whole new model, and x402, Nanopayments, and Tempo are building that model.

As Citrini's constructed model suggests, any disruption to consumer commerce will likely occur later. It requires agents to handle a significant share of discretionary spending, which in turn requires consumers to trust agents to make purchasing decisions they currently make themselves.

Visa is being pressured by higher-quality customers. These customers no longer need the elements that Visa relies on for success. The 2-3% exchange fee is not a transaction tax; it is a tax on the irrational behavior of humanity. And agents are entirely rational.

How do I know this is important? Because Visa spent $1.8 billion last week to ensure it is not excluded from the answers.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。