On June 16, 2026, Adyen introduced “Adyen Agentic” in New York, a three-layer API suite designed to sit between merchants and the burgeoning ecosystem of AI commerce agents. The architecture is modular: Agentic Feed handles product discovery and catalogs, Agentic Cart manages checkout orchestration, and Agentic Payments covers authentication, tokenization, fraud, and settlement. The company positions this as a “universal translator,” promising that merchants can integrate once and remain compatible with Google UCP, Google AP2, and OpenAI’s ACP simultaneously. It is a pitch for simplicity in an increasingly fragmented landscape.
For enterprise retailers, the current reality of agentic commerce is a repetitive engineering burden. Every new agentic surface — whether a chatbot, a voice assistant, or a specialized shopping agent — demands that merchants rebuild their integration logic from scratch. Karan Katyal, Adyen’s Global Head of Agentic Commerce, framed the problem clearly:
Every new agentic surface asks merchants to rebuild from scratch. We believe the future of agentic commerce should be open, so we intentionally designed Adyen Agentic to help retailers integrate once and participate across evolving platforms, protocols, and experiences — without having to bet on which ecosystems ultimately win.
The technical mechanism here is protocol-agnosticism. By abstracting the complexity of different agent protocols into a single merchant-side layer, Adyen is attempting to decouple the merchant’s backend from the specific “war” being fought between competing standards like Google’s UCP and OpenAI’s ACP. This approach relies on the assumption that payment processors are structurally incentivized to remain neutral. Because processors earn revenue based on transaction volume, they are indifferent to which protocol facilitates the sale, provided the transaction clears. This creates a natural alignment between the processor’s profit motive and the merchant’s desire for a single, stable integration point.
This strategy stands in stark contrast to Stripe’s approach. Stripe has tied its agentic efforts closely to OpenAI and the ChatGPT ecosystem, focusing on a buyer-side, protocol-coupled model. While Stripe’s ACP integration is deeply embedded within the ChatGPT experience, Adyen is positioning itself as the infrastructure layer that sits behind the merchant, regardless of which agent the consumer chooses to use. It is a battle of philosophies: Stripe is betting on the dominance of specific buyer-side interfaces, while Adyen is betting on the persistence of the merchant’s need for a unified, platform-agnostic payment gateway.
However, the “universal translator” thesis faces a significant structural hurdle: the consumer trust gap. According to data from Product.ai (April 2026), only 14% of US online shoppers trust AI recommendations without verification, and 42% refuse to trust AI for purchases exceeding $25. Even with robust infrastructure, 86% of consumers still verify their purchases before finalizing them. No amount of protocol-agnostic middleware can solve the fact that, for now, the human remains the final, skeptical arbiter of the transaction. Infrastructure can facilitate the flow of data, but it cannot manufacture the trust required to move high-value commerce into fully autonomous agentic loops.
Furthermore, the current state of Adyen Agentic is characterized by limited availability for US enterprise merchants, and the company has disclosed no adoption metrics. While early partners like American Express, Mastercard, Salesforce, and Visa, along with retailers such as ESW, Scheels, Sézane, and SharkNinja, have signed on, the actual volume flowing through these rails remains unproven. As Nicolas Benoist, CTO at Sézane, noted:
With Adyen Agentic, we were able to build on proven, battle-tested foundations to deploy quickly and confidently. As the agentic ecosystem grows, the solution will allow us to seamlessly reach consumers across every emerging commerce platform.
The current protocol wars pit three distinct architectures against one another: x402, which handles high-volume but low-value crypto-native stablecoin settlements; Google UCP, which leverages existing card-rail checkouts; and Mastercard’s AP4M, a multi-rail approach incorporating stablecoins. The industry assumes one must emerge as the dominant standard, but this binary view ignores the role of middleware. If merchants can successfully offload the technical complexity of these protocols to providers like Adyen, the competitive advantage shifts away from the underlying rails and toward whoever controls the merchant interface. In this scenario, the protocol itself becomes a commodity, and the middleware provider becomes the gatekeeper of the merchant’s agentic strategy.
Adyen’s model is currently limited to a handful of US enterprise merchants with no public adoption metrics to validate the scale. Meanwhile, consumer behavior remains stubbornly cautious: only 14% of users trust AI recommendations without manual verification, and 42% refuse to use AI for purchases exceeding $25. The infrastructure is being built, but the transaction volumes required to justify it are absent. For now, the industry is busy optimizing the plumbing for a house that hasn’t been built yet.