Inside Adyen Agentic: The Universal API Layer for Multi-Protocol AI Commerce

Agentic Feed is the structured product and inventory layer. It makes a merchant’s catalog data available in a format that conversational AI platforms can surface during natural-language shopping interactions . This solves one of the earliest friction points in agentic commerce: AI assistants need machine-readable product data to suggest items, check availability, and understand attributes like sizing, color, or compatibility—but most merchant catalogs are not natively optimized for conversational discovery.

Agentic Cart acts as the orchestration layer, bridging a merchant’s existing commerce and checkout infrastructure to the various conversational commerce surfaces . Rather than rebuilding cart logic for each AI platform, merchants use Agentic Cart to map their current order flows—including promotions, taxes, loyalty logic, and fulfillment rules—into whatever protocol a given platform expects. This is the connective tissue that Adyen argues is missing from earlier point-to-point integrations.

Agentic Payments is the transactions and risk layer for agent-led purchases. It handles secure payment processing for AI-commerce flows, building on Adyen’s existing single-platform infrastructure that already spans gateway, risk management, processing, acquiring, and settlement . The company processes over 1.4 trillion euros in annual transaction volume, giving this layer a foundation that predates the current agentic-commerce hype .

Critically, the three layers are designed to be modular: a merchant might use Agentic Feed with an existing cart system, or adopt Agentic Payments while handling discovery elsewhere . Adyen’s pitch is that this flexibility lets retailers participate in the agentic-commerce shift incrementally, rather than requiring a wholesale replatforming.

Strategic Partners and Ecosystem Support

Adyen Agentic’s launch was accompanied by a roster of high-profile payment and commerce companies signaling support. Early participants and ecosystem backers named in connection with the announcement include Mastercard, Visa, American Express, and Salesforce .

Mastercard’s involvement extends beyond the Adyen-specific launch. In June 2026, Mastercard separately unveiled its Agent Pay ecosystem, which includes more than 30 industry leaders—among them Adyen, Stripe, Checkout.com, Coinbase, and Global Payments—as early adopters and supporters . Sherri Haymond, EVP and Global Head of Digital Commercialization at Mastercard, framed trust, security, and merchant enablement as non-negotiable requirements for the category to scale :

"For agentic commerce to scale, trust, security and interoperability need to be built in from the start. Equally critical is enabling merchants — making data discoverable and machine‑readable so they can participate as models and protocols evolve. With Agent Pay, Mastercard is helping establish a strong foundation for secure, scalable agent‑driven transactions, while solutions like Adyen Agentic are contributing to a more vibrant, connected ecosystem."

Visa and American Express serve as long-standing payment network partners, while Salesforce’s inclusion signals reach into the enterprise commerce-cloud segment, where Adyen already maintains a technology partnership . While Adyen’s announcement also mentioned merchant partners including ESW, Scheels, Sézane, and SharkNinja, available public sources focus primarily on payment and technology ecosystem partners at launch .

Which Protocols Does Adyen Agentic Build On?

Adyen’s core strategic positioning with Agentic is its explicit multi-protocol compatibility. Rather than aligning with a single technical standard, Adyen Agentic is designed to work across the major protocols currently competing to define agentic-commerce infrastructure .

Meta’s AI checkout is identified as compatible from launch, with additional platforms expected over time . This gives Adyen Agentic immediate relevance for merchants wanting to sell through Meta’s conversational AI surfaces.

Google’s Universal Commerce Protocol (UCP) is perhaps the most ambitious of the protocols Adyen supports. Announced at NRF in January 2026 and co-developed with companies including Shopify, Etsy, Wayfair, Target, and Walmart, UCP defines the full commerce workflow—from product discovery through checkout and order management—as a common language for AI agents and merchant systems . Over 20 partners endorsed it at launch, including Adyen, American Express, Stripe, and Visa . UCP powers direct buying through AI Mode in Google Search and Gemini .

The Agent Payments Protocol (AP2) sits one layer down in the stack as the payment-specific standard. Developed by Google with more than 60 partners, AP2 provides a framework for AI agents to initiate financial transactions on a user’s behalf using cryptographically signed mandates—without the agent ever accessing real banking credentials . AP2 can be used as an extension of Google’s broader Agent2Agent protocol, making it a key piece of the trust and authorization infrastructure for delegated AI purchases .

OpenAI’s Agentic Commerce Protocol (ACP) is maintained by OpenAI and Stripe, first announced in September 2025 and deployed inside ChatGPT’s Instant Checkout . ACP is built around a Shared Payment Token model: a user’s payment method is stored in ChatGPT, and when a purchase is confirmed, Stripe issues a single-use credential scoped to that specific merchant and cart total . Adyen Agentic is explicitly positioned to support ACP alongside UCP and AP2, reinforcing its multi-protocol stance rather than treating any one standard as the winning horse .

Together, these three protocols—ACP, UCP, and AP2—represent the main technical standards defining how AI agents transact with merchants on behalf of humans, though they serve overlapping but distinct layers of the commerce stack .

The Fragmentation Problem Adyen Claims to Solve

Karan Katyal, Global Head of Agentic Commerce at Adyen, articulated the core problem the product addresses in terms that resonate with any enterprise CTO facing the current AI-platform landscape :

"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 practical reality he describes is this: an AI checkout experience on Meta’s surfaces can require different integration patterns, catalog schemas, cart logic, and payment orchestration than a purchase flowing through Google’s Gemini or OpenAI’s ChatGPT. For a large retailer, supporting all three natively could mean maintaining three separate technical integrations—each with its own data synchronization, error handling, and ongoing maintenance burden. Adyen Agentic aims to collapse that complexity into a single integration that translates a merchant’s systems into whatever protocol a given AI platform requires .

This is fundamentally a bet on interoperability as a service. Adyen is wagering that enterprises will prefer to pay a middleware layer that abstracts away protocol differences, rather than investing in dedicated engineering teams for each new AI sales channel as it emerges.

How This Positions Adyen Against Stripe

The launch crystallizes a philosophical divide in the emerging agentic-commerce infrastructure market.

Stripe is deeply associated with the OpenAI-linked ACP framework. The company co-developed the protocol, powers the payment rails under ChatGPT’s Instant Checkout, and has built its agentic-commerce narrative around that close platform alignment . Stripe’s position is that being the payment engine inside the most prominent consumer-facing AI assistant is the most direct path to volume.

Adyen is taking the opposite stance. By explicitly supporting ACP, UCP, AP2, and Meta’s AI checkout simultaneously, Adyen Agentic positions Adyen as a neutral, merchant-first infrastructure layer in a market where platform alliances are hardening . The argument is twofold: first, that no single AI platform or protocol will dominate agentic commerce in the way that a single search engine or social network might dominate its category; and second, that merchants will bristle at being forced to pick sides between their biggest technology partners.

There is some evidence that the market structure supports Adyen’s view. Mastercard’s Agent Pay ecosystem includes both Adyen and Stripe among its early participants, suggesting that even the card networks see value in a multi-rail, multi-protocol approach . Meanwhile, major merchants like Etsy and Shopify have engaged with multiple protocols rather than committing exclusively to one .

Adyen’s existing infrastructure also gives it a foundation that is independent of any single AI protocol. The company’s single-platform architecture already processes over 1.4 trillion euros in annual volume across gateway, risk, processing, acquiring, and settlement—and its developer resources describe Agentic Commerce as an extension that enables AI agents to discover products and process secure payments using that same infrastructure .

In practice, the competitive dynamic is likely to play out along customer-segment lines. Retailers deeply embedded in the OpenAI ecosystem or building ChatGPT-native shopping experiences may gravitate toward Stripe’s tighter ACP integration. Enterprise retailers that sell across multiple platforms—and that value independence from any single Big Tech partner—are the natural audience for Adyen’s multi-protocol pitch. Adyen is betting that the latter group is large enough, and wary enough of platform lock-in, to make a universal translation layer the smarter infrastructure bet .