Digital Money Has a New Payment Standard and It’s Not Built For Humans

On Thursday (April 2), those legacy payments got an autonomous upgrade with the transfer of the x402 protocol to the Linux Foundation, a nonprofit organization focused on open source. At the center of this initiative is an unlikely candidate: HTTP status code 402, “Payment Required,” a relic of the early internet that was never meaningfully implemented.

Now, under what’s being called the x402 standard, that dormant code is being revived as a native mechanism for enabling micropayments directly within web requests. Instead of redirecting users to checkout flows or embedding third-party APIs, the payment becomes part of the protocol itself as something executed automatically, invisibly and continuously.

The catch? The payments relying on this new protocol use stablecoins, and they’re not mean for humans. They’re designed for artificial intelligence bots.

And x402 isn’t the only protocol out there trying to impose a new standard on the emerging agentic marketplace.

See also: What Happens to Stores When AI Agents Do the Shopping? 

Scaling Payments Without Interfaces

The web has a native protocol for information exchange, HTTP, but no equivalent for value exchange. Payments are bolted on through intermediaries: card networks, payment gateways, and subscription models. These systems are optimized for human decision-making, not machine autonomy.

To understand the new payment protocol, it can help to contrast x402 with those consumer-facing systems. Platforms like Apple Pay and PayPal abstract complexity behind polished interfaces, prioritizing trust, ease and regulatory compliance. They are optimized for discrete, intentional transactions such as buying a product, sending money, subscribing to a service.

The x402 protocol, by contrast, removes the interface entirely. There is no checkout page, no user authentication in the traditional sense. Payments are executed programmatically, often by autonomous agents — bots, services or AI systems — that negotiate access to resources on the fly.

In theory, this collapses authentication, billing and access into a single interaction. The result is a model where services can be priced per request, and agents can transact continuously without pre-negotiated accounts.

Early participants in the x402 Foundation include Adyen, Amazon Web Services, American Express, Ampersend.ai, Base, Circle, Cloudflare, Coinbase, Fiserv Merchant Solutions, Google, Kakao Pay, Mastercard, Merit Systems, Microsoft, Polygon Labs, PPRO, Shopify, Sierra, Solana Foundation, Stripe, thirdweb and Visa.

See also: Why Fortune 500 Firms Are Becoming Blockchain Validators 

A Fragmented Stack, Not a Single Solution

Yet x402 is only one piece of a broader puzzle. The emerging ecosystem of agentic payment protocols is less a winner-take-all race than a layered architecture, with different players asserting control over different functions.

At the base is payment execution, where x402 operates. It defines how money moves between parties in a machine-native way. But it assumes that the agent already has access to funds and the authority to spend them.

That authority is the domain of protocols like Google’s Agents-to-Payments (AP2). AP2 focuses on governance: who authorizes an agent, what limits apply, and how transactions are audited. In enterprise contexts, this layer is indispensable. Companies are unlikely to grant autonomous systems unrestricted spending power without enforceable policies and traceability.

Above that sits the commerce layer, where systems such as Stripe’s Agent Commerce Protocol (ACP) aim to adapt existing payment infrastructure for AI-driven transactions. ACP enables agents to interact with merchants using familiar rails — credit cards, payment gateways and checkout flows — while abstracting the complexity behind APIs.

These layers are complementary rather than mutually exclusive. A single transaction could involve AP2 authorizing the spend, ACP facilitating the purchase, and x402 executing the payment. The real competition is not between protocols in isolation, but between competing visions of how tightly these layers should be integrated and who should control them.

The PYMNTS Intelligence report “How Acquirers Prepare for Agentic Commerce” found that nearly 80% of surveyed acquirers said they are at least somewhat prepared to support seamless omnichannel shopping experiences, a prerequisite for any system in which autonomous agents transact across digital and physical environments.

Read more: Stablecoin Plans Split as Banks Go Their Own Way

Competing Visions of Machine Money

The divergence between the emerging approaches reflects deeper strategic fault lines across the agentic commerce landscape. Coinbase’s x402 represents a crypto-native vision. It assumes that the most efficient way for machines to transact is through push-based payments using stablecoins, bypassing traditional intermediaries. In this model, agents hold funds directly and send them as needed, much like servers exchange data packets. The emphasis is on programmability, composability, and minimal friction.

By contrast, AP2 and related efforts from large technology firms reflect an enterprise-centric worldview. Here, the priority is not frictionless payment but controlled delegation. Agents are extensions of organizations, operating within predefined budgets and policies. Payments may still rely on existing financial systems, but are wrapped in layers of governance and compliance.

And, ultimately, despite the conceptual clarity of these approaches, significant obstacles remain. Adoption is the most immediate hurdle. Protocols derive value from network effects, and the history of the internet is littered with technically elegant standards that failed to gain traction.

User experience is another constraint. While agents may handle transactions autonomously, humans remain accountable for their behavior. Trust in systems that can spend money without explicit approval is not guaranteed. Mechanisms for oversight, error handling, and dispute resolution will be critical.

Finally, there is the question of economic viability. Micropayments have long been touted as a solution to digital monetization but have struggled to achieve widespread adoption. Even if technical barriers are removed, behavioral and market dynamics may limit their impact.