Who Pays the Agent? The Race for Frictionless Machine Payments

The Protocols Are Not Competing; They Are Stacking

In September 2024, if you wanted an AI agent to pay for something, you had essentially one, unsafe option. Twelve months later, four architectures exist, backed by some of the largest companies in technology.

Coinbase built x402, a crypto-native protocol that turns stablecoin wallets into universal API keys. Stripe and Tempo launched MPP, a payment-method-agnostic standard that handles cards, crypto, and Lightning through a single HTTP flow. Google assembled AP2, an authorisation layer that lets users delegate spending power to agents via cryptographic mandates. And Visa extended its existing card rails to provision AI-ready tokenised credentials.

What most coverage misses is that these four approaches aren’t purely competing. There’s genuine overlap at the protocol layer, but the more important dynamic is that they’re assembling into a stack. We believe the right question isn’t “which protocol wins?” but “which companies capture the most layers, and therefore attract the most value?”

The $0.30 Wall

Across 176 million x402 payments to date, the median transaction sits between $0.01 and $0.10, with 76% of activity falling below the $0.30 card-fee floor. That number tells you almost everything about why traditional payment rails can’t serve this market. A fixed processing fee of roughly 30 cents per transaction makes sub-dollar payments uneconomical. An agent paying three cents for a weather API call can’t route through Visa.

Layer 2 stablecoin settlement costs $0.0001. For agents, this means blockchain rails are a necessity. 

A Single Stablecoin Dominates

Of those 176 million payments, 98.6% settled in USDC. Stablecoins won the settlement layer for machine commerce almost by default; they were the only instrument that could handle sub-dollar transactions without the economics collapsing.

That concentration is both a validation and a vulnerability. It validates Circle’s positioning as the default settlement asset, but it also means the entire agent payments ecosystem depends on a single stablecoin issuer’s reserve management, regulatory standing, and technical infrastructure. Nobody in the space is publicly discussing this. We think they should be.

The Vertical Integration Race

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

Over the past twelve months, incumbents have poured over $8 billion into acquisitions to fill gaps in their stack coverage. Capital One paid $5.15 billion for Brex, Mastercard spent $1.8 billion on BVNK, and Stripe acquired Bridge. These are infrastructure consolidation plays by companies that see machine payments as an inevitable expansion of their core business.

From Bot Activity to Agent Commerce 

The machine economy is already here. It just isn’t doing commerce yet. But the signals are clear: AI agents account for 37% of all Safe transactions on Gnosis Chain, rising to over 75% on peak days. Coinbase has deployed tens of thousands of agents with built-in guardrails. 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 what makes it possible.

Regulation Is the Binding Constraint

MiCA, the GENIUS Act, and the EU AI Act all reach enforcement within weeks of each other in mid-2026. None of them address autonomous machine-to-machine transactions. This isn’t a future problem; it’s a current one, unfolding on a live timeline with real capital at stake.

What Comes Next

The market is moving toward greater agent autonomy, but we don’t think the pace will be set by the technology, which is largely ready. It’ll be set by the trust infrastructure that makes it safe. The fully permissionless vision is intellectually appealing, but it assumes a level of AI reliability that doesn’t exist yet. Until agents stop hallucinating, they probably shouldn’t have unsupervised access to a user’s funds.

We think the bottom-up thesis is the most compelling framing for what happens next. Crypto rails have already won micropayments by default. As volume builds and trust infrastructure matures, progressively larger transaction values will migrate onchain. The question isn’t whether machine-native payments will scale, but how quickly the trust layer catches up to the settlement layer.

Get the full insights. Download the report today. 

This article highlights the core findings of our research. The full report explores the data in depth, including protocol architecture analysis, interview insights from Coinbase and Tempo, transaction economics modelling, and the regulatory landscape.