As artificial intelligence agents move from experimental demos into real-world commerce, they face a fundamental problem: they cannot open bank accounts. This structural limitation is forcing major technology and payments companies to build crypto-native infrastructure for machine-to-machine transactions, reshaping how Web3 settlement layers fit into enterprise commerce.

The emerging "agentic commerce" market could reach between $3 trillion and $5 trillion by 2030, according to McKinsey estimates. Unlike human-driven commerce, where credit cards and traditional banking work well, autonomous AI agents require programmable, always-on payment systems that operate without human intermediaries. This gap is creating a new infrastructure race among incumbents and specialized blockchain builders.

What Are the Three Layers of Agentic Commerce Infrastructure?

Industry experts and companies are organizing the buildout around three distinct infrastructure layers, each requiring different technical approaches and serving different parts of the transaction flow:

• Compute and Models Layer: The AI platforms and reasoning engines that enable agents to understand and act on information, with companies like Nvidia leading development of enterprise-grade agent toolkits.
• Payment and Identity Rails Layer: The systems that process transactions and verify participants, where traditional payment networks like Visa and PayPal are integrating stablecoin settlement alongside card networks.
• Settlement Layer: The underlying blockchain infrastructure that actually moves value between parties, where smaller, infrastructure-focused Web3 companies are positioning themselves as connective layers between existing networks.

Each layer attracts different players. Nvidia is building the platform layer for enterprise AI agents. Visa and PayPal are wiring stablecoins, which are cryptocurrencies pegged to stable assets like the US dollar, into their existing payment infrastructure. Smaller blockchain companies are targeting the cross-chain settlement layer beneath it all.

How Are Major Payments Companies Positioning Themselves?

Visa reported a $7 billion annualized stablecoin settlement run rate in its fiscal Q2 2026 earnings, up more than 50% from the prior quarter. The company's stablecoin-linked card payment volume rose nearly 200% year-over-year, signaling rapid adoption of on-chain settlement for real transactions. Visa now operates across nine blockchains, including Polygon, Base, and Circle's Arc, and serves as a validator on multiple blockchain networks, meaning it helps govern parts of the settlement infrastructure itself.

"A hyperscaling bridge layer between stablecoin and real-world solutions and applications for users," is how Visa CEO Ryan McInerney described the company's strategy, with unit economics similar to its existing card business.

Ryan McInerney, CEO at Visa

PayPal has deployed its PYUSD stablecoin across 13 blockchains and launched an open-source PayPal Agent Toolkit that exposes its APIs to large language models. The company joined a 120-partner consortium called the Agentic Payments Protocol (AP2), led by Google, to standardize how AI agents interact with payment systems.

"AI agents structurally cannot use traditional bank accounts, making crypto-native payment rails a necessary layer for autonomous commerce," explained Richard Widmann, Web3 strategy head at Google Cloud, at Consensus Miami in May.

Richard Widmann, Web3 Strategy Head at Google Cloud

PayPal's positioning is straightforward: if AI agents become a meaningful commerce layer, programmable crypto rails become a natural payment infrastructure for those systems.

What Role Are Smaller Infrastructure Players Taking?

While Nvidia, Visa, and PayPal dominate headlines, smaller infrastructure-focused companies are targeting the cross-chain settlement layer. The Crypto Company (OTC: CRCW) is building Frame, a Layer 1 blockchain designed to simplify cross-chain transactions and support AI-driven commerce use cases. Rather than competing directly with established networks like Ethereum or Solana, Frame is positioned as a connective layer that links existing blockchain networks and routes liquidity across them.

Frame's revenue model depends on network usage, including transaction fees and infrastructure services such as RPC (remote procedure call) providers, oracles, and relay functions. RPC providers allow applications to communicate with blockchain networks; oracles feed external data into smart contracts; relay functions route transactions across chains. The company committed $2 million to Frame development and eliminated roughly $4 million in legacy convertible debt in late 2025, targeting a 2026 mainnet launch.

For investors and developers monitoring this space, the signal from incumbents is clear: the largest payments network on the planet is treating on-chain settlement and AI agent commerce as core to its next decade of growth. This institutional commitment suggests that Web3 infrastructure is no longer a niche concern but a foundational layer for the next generation of commerce.

Why Does This Matter for Web3 Infrastructure Providers?

The convergence of AI agents and crypto payment rails creates a structural demand for Web3 infrastructure that did not exist before. Nvidia's announcement of 17 enterprise software companies planning to use its Agent Toolkit, including Adobe, Salesforce, SAP, Palantir, and Siemens, signals that agentic AI is moving into production at scale. If enterprise AI agents proliferate as industry leaders forecast, demand for machine-native payment and settlement infrastructure could expand alongside them.

For Web3 infrastructure providers, this represents a shift from speculative token markets to utility-driven adoption. The question is no longer whether blockchain settlement will be used for commerce, but how quickly it will scale and which infrastructure layers will capture the most value.