Machines Pay, Humans Reap: Coinbase, Stripe, Google, and Visa Battle for AI-Payment Dominance

Author: Ben Harvey

Translated and compiled by TechFlow

TechFlow Introduction: A year ago, machine-to-machine (M2M) payments were merely a concept. Today, Coinbase, Stripe, Google, and Visa have each deployed competing payment architectures. AI Agents have already executed 176 million transactions and settled $73 million—and traditional financial giants have spent over $8 billion on acquisitions to secure positions in this emerging payment stack. This is not a futuristic narrative; it is the ongoing reorganization of payment infrastructure—where value accrues to those who control the most layers.

A year ago, machine-to-machine payments were still just a concept. Today, four competing payment architectures are live—backed by Coinbase, Stripe, Google, Visa, and American Express. AI Agents have settled over $73 million across 176 million transactions, and traditional incumbents have invested more than $8 billion in acquisitions to stake their claim in this new payment stack.

This report—produced jointly by Keyrock, Coinbase, and Tempo—examines how this payment stack is assembled, whether its economic model holds, and what obstacles lie ahead.

Protocols Aren’t Competing—They’re Stacking

In September 2024, if you wanted an AI Agent to make a payment, you had essentially one insecure option. Twelve months later, four distinct architectures are live—backed by some of the largest companies in tech.

Coinbase built x402, a crypto-native protocol that turns stablecoin wallets into universal API keys. Stripe and Tempo launched MPP (Machine Payment Protocol), a payment-method-agnostic standard that processes card payments, cryptocurrencies, and Lightning Network transactions through a single HTTP flow. Google assembled AP2, an authorization layer enabling users to delegate payment permissions to Agents via cryptographic consent. Visa extended its existing card rails to offer AI-ready tokenized credentials.

What most coverage overlooks is that these four solutions aren’t purely competitive. While there is overlap at the protocol layer, the more important dynamic is that they are being assembled into a unified payment stack. The right question isn’t “Which protocol will win?” but rather “Which companies control the most layers—and therefore capture the most value?”

The $0.30 Wall

Of the 176 million x402 payments processed to date, the median transaction size falls between $0.01 and $0.10—and 76% of activity occurs below the $0.30 interchange fee floor for card networks. That number almost entirely explains why traditional payment rails cannot serve this market. A fixed processing fee of roughly $0.30 per transaction renders microtransactions unprofitable. An Agent paying $0.03 for a weather API call simply cannot route that through Visa.

L2 stablecoin settlement costs $0.0001 per transaction. For Agents, this means blockchain rails are essential.

Single Stablecoin Dominance

Of those 176 million payments, 98.6% were settled in USDC. Stablecoins have effectively won the settlement layer for machine commerce by default—they are the only instruments capable of handling microtransactions without collapsing the underlying economics.

This concentration represents both validation and vulnerability. It validates Circle’s position as the default settlement asset—but also means the entire Agent payment ecosystem depends on a single stablecoin issuer’s reserve management, regulatory standing, and technical infrastructure. No one in the industry is publicly discussing this. We believe they should.

The Vertical Integration Race

Coinbase and Stripe each cover five of the six layers in the emerging payment stack. Coinbase controls the settlement layer (Base), wallets (Agentic Wallets), routing (internal infrastructure), the payment protocol (x402), and governance (as an AP2 partner). Stripe mirrors this with Tempo (settlement), Privy (wallets), Bridge (routing—acquired for $1.1 billion), MPP (protocol), and its compliance infrastructure.

Over the past 12 months, traditional incumbents have spent over $8 billion on acquisitions to fill gaps in their payment stack coverage. Capital One acquired Brex for $5.15 billion; Mastercard bought BVNK for $1.8 billion; Stripe acquired Bridge. These are infrastructure consolidation moves—by firms treating machine payments as an inevitable core business expansion.

From Bot Activity to Agent Commerce

The machine economy has arrived—though it hasn’t yet begun doing commerce. But the signals are clear: AI Agents account for 37% of all Safe transactions on Gnosis Chain—and at peak times, over 75%. Coinbase has deployed tens of thousands of Agents with built-in safeguards. Over 104,000 Agents are registered across 15 or more directories and registries.

The shift from extractive bot activity to productive Agent commerce is underway. The payment infrastructure examined in this report is precisely what makes that transition possible.

Regulation Is a Constraint

MiCA, the GENIUS Act, and the EU AI Act are all scheduled to enter enforcement phases within weeks of each other in mid-2026. None of them address autonomous machine-to-machine transactions. This is not a future problem—it is a present one, unfolding in real time against a backdrop of real capital commitments.

What Comes Next

The market is trending toward greater Agent autonomy—but we believe the pace will not be set by technology, which is largely ready. Instead, it will be dictated by the trust infrastructure required to make it safe. A fully permissionless vision is theoretically appealing—but it assumes an AI reliability level that does not yet exist. Until Agents stop hallucinating, they probably shouldn’t have unsupervised access to user funds.

We believe the bottom-up argument offers the most compelling framework for what comes next. Crypto rails have already won microtransactions by default. As transaction volume grows and trust infrastructure matures, ever-larger transaction values will migrate on-chain. The question is not whether machine-native payments can scale—but how quickly the trust layer can catch up to the settlement layer.

This article summarizes the report’s key findings. The full report dives deeper into the data—including protocol architecture analysis, insights from interviews with Coinbase and Tempo, transaction economics modeling, and the regulatory landscape.