Google Adds UCP Checkout Support to Merchant API as Agentic Commerce Develops

Google is expanding the infrastructure behind agentic commerce, adding support for Universal Commerce Protocol (UCP) checkout to its Merchant Accounts API.

According to Google’s latest Merchant Accounts API release notes, the August update introduced ucp_checkout_management as a new account service type. Google also added support for the ucp-integration program and corresponding UCP integration terms of service. The features currently have limited access.

Although the update is technical, it points toward a larger change in ecommerce: AI systems are increasingly being designed not only to help consumers discover products, but also to interact with the systems required to complete commerce workflows.

For retailers, brands and ecommerce technology providers, this makes the infrastructure behind product discovery increasingly important.

Google Extends Its Merchant Infrastructure

Google’s Merchant API provides businesses and commerce platforms with programmatic access to Merchant Center functionality. It can be used to manage products, accounts, data sources and other elements involved in presenting merchant information across Google.

The Accounts API specifically handles Merchant Center accounts and relationships between merchants and service providers.

Google’s August update adds UCP-related capabilities to this account infrastructure. The release notes list three additions: UCP checkout management as a service type, support for a UCP integration program and a dedicated terms-of-service category.

Google stresses that UCP checkout currently has limited access, so this should not be interpreted as a general rollout to every Merchant Center user.

Nevertheless, its appearance in the Merchant API further indicates that Google is preparing its commerce infrastructure for more agent-driven interactions.

From Product Discovery Toward Transactions

AI has already changed how consumers search for products.

Instead of entering a few keywords and browsing pages of results, shoppers can increasingly describe an intention: find a television within a particular budget, compare laptops for specific work requirements or suggest products that meet a detailed list of preferences.

Agentic commerce takes that idea further.

An AI system can potentially perform more of the steps between a consumer expressing an intention and completing a transaction. That requires connections between AI interfaces and merchant systems rather than simply generating recommendations in a conversation.

Checkout is therefore an important part of the infrastructure.

Google’s addition of UCP-related services to its Merchant API suggests that merchant accounts, permissions and commerce integrations are being prepared for this type of interaction.

Agents Still Need to Understand the Product

As you move closer to transactions, product information quality matters more, not less.

Before an AI system can help a customer purchase a product, it must first identify which product satisfies the customer’s request.

Consider a shopper asking an AI assistant for a 55-inch television with a particular display technology, connectivity options and energy characteristics. The agent needs reliable information about those attributes before it can compare suitable products.

The same applies across ecommerce categories. Compatibility, dimensions, materials, technical specifications, variants, identifiers and other characteristics can determine whether a recommendation is useful or incorrect.

This is where structured product data becomes particularly relevant to Icecat’s ecosystem.

Icecat standardizes product information so it can be exchanged and reused across brands, distributors, retailers and other e-commerce channels. In an agentic environment, the same structured information can also help machines interpret catalogs and distinguish between products.

The interface may increasingly become conversational, but underneath it remains a data problem: the system needs accurate information in a format it can understand.

Google’s Merchant APIs Are Becoming Richer

The UCP update also sits within a wider expansion of Google’s Merchant API.

During 2026, Google has introduced additional product-related capabilities covering areas such as video links, return policies and conversational product attributes. In May, Google added fields including questions and answers, popularity rank, document links, variant options and related products.

More recently, the Products API added links for certification documents and certification labels.

Taken together, these additions indicate that merchant data is becoming richer than the traditional combination of title, image, price and availability.

That matters as AI takes a larger role in discovery. A search engine can rank a page partly from keywords and other signals. An agent answering a detailed product request needs enough structured context to reason about which products meet the requirements.

For merchants, completeness and consistency across product attributes may therefore become increasingly important for visibility in AI-mediated shopping environments.

Product Data Becomes Infrastructure for Agentic Commerce

Agentic commerce is still developing, and Google’s UCP checkout functionality remains limited in availability. It is therefore too early to know how quickly these systems will change everyday shopping behavior.

But the technical direction is becoming clearer.

AI interfaces are moving closer to merchant infrastructure, while commerce APIs are evolving to support richer interactions between platforms, merchants and automated systems.

For e-commerce businesses, preparing for that environment isn’t just about adopting another AI tool. It also means ensuring that the underlying catalog can be reliably interpreted by those tools.

Accurate identifiers, standardized specifications, rich attributes, documentation and consistent product relationships provide the foundation from which AI agents can search, compare and ultimately act.

As checkout becomes more accessible to AI-driven commerce systems, the gap between product data and the transaction shrinks.

For brands and retailers, that makes the quality of the information describing a product an increasingly important part of whether an AI agent can find it, understand it and confidently bring it into a customer’s purchasing journey.