This week, Google announced the Universal Commerce Protocol (UCP), an open-source specification designed to enable agent-driven commerce experiences in the era of AI. UCP defines how AI systems can interact with merchants across the shopping journey, from product discovery through checkout and ordering, primarily within Google-controlled surfaces.

This follows OpenAI’s earlier introduction of the Agentic Commerce Protocol (ACP), its own attempt at establishing a common framework for commerce actions executed directly by AI agents. ACP focuses on standardizing how agents discover products, evaluate options, and complete transactions across the open web.

Together, these announcements signal a shift toward AI systems taking on a more active role in commerce. We are moving toward an era where AI agents not only assist users during search, comparison, and checkout, but can also act on expressed intent and complete purchases autonomously on their behalf.

In this article, we’ll look at how UCP and ACP differ in scope and design, what each protocol enables today, and how D2C brands can prepare for a future where AI-led shopping becomes an additional, increasingly important commerce channel.

TMO provides end-to-end eCommerce services from platform optimization to system integrations, Agentic AI, and scalable commerce architectures.

The Common Goal of Agentic Commerce

Agentic Commerce refers to commerce interactions where AI agents act on behalf of users across parts of the shopping journey such as:

• Surfacing product information
• Comparing options
• Validating availability or pricing
• Selecting products
• Initiating checkout
• Completing transactions

For AI agents to interact with merchant systems in a predictable and reliable way, commerce infrastructure needs to be legible, structured, and safe for machine-driven use. This is where standardization becomes necessary, and what both Google and OpenAI are attempting to address by defining a open-source common language for agent-led commerce.

OpenAI's Agentic Commerce Protocol (ACP)

Co-developed by OpenAI and Stripe, the Agentic Commerce Protocol (ACP) is an open standard that enables programmatic exchanges between buyers, AI agents, merchants, and payment providers to complete a purchase. ACP requires merchants to support three core flows:

• Product Feed: A merchant’s live, structured product data including descriptions, media, pricing, availability, and fulfillment options, allowing agents to accurately surface products to users.
• Agentic Checkout: A standardized specification that allows agents to render checkout directly within their own interfaces, while keeping order management, payments, and compliance on the merchant’s existing commerce stack.
• Delegated Payment: A secure mechanism that lets agents share payment details with merchants and payment service providers, who then process transactions as they would for any other order.

While ACP is open for any merchant to implement. Instant Checkout, the first solution built on the protocol, is currently available to Etsy and Shopify sellers in the US for ChatGPT across web, iOS, and Android. Support for multi-item carts, additional merchants, and more regions is expected to follow.

Google's Universal Commerce Protocol (UCP)

Similarly, UCP was co-developed by Google alongside industry partners including Shopify, Etsy, Wayfair, Target, and Walmart. It aims to define a common language and a set of functional primitives that standardize the commerce journey, from discovery and evaluation through purchase and order management, into a single integration layer.

UCP is designed to work with existing industry protocols such as Agent2Agent (A2A), Agent Payments Protocol (AP2), and Model Context Protocol (MCP), and currently focuses on three core capabilities:

• Checkout: Enables platforms to offer native or UI-embedded checkout experiences that users can complete manually. UCP supports customization and complex logic such as dynamic pricing and tax calculation, while retailers remain the Merchant of Record and continue to use their existing commerce infrastructure.
• Identity Linking: Uses OAuth 2.0 standards to allow agentic services to securely obtain authorization to perform actions on behalf of users, without sharing credentials.
• Order: Provides access through webhooks to a complete record of transactions, including Line Items, Fulfillment, and Adjustments data, powering order confirmations, status updates, shipment tracking, and return processing.

The Checkout capability is will roll out to Google’s AI Mode in Search and the Gemini app for US retailers. Google Pay is currently supported, with PayPal announced as in plans. Additional capabilities such as product discovery, cart building, loyalty programs, and native cross-sell and upsell modules are outlined on UCP’s roadmap.

Alongside the UCP announcement, Google also introduced Business Agent, a chatbot experience that allows retailers to interact with users directly within Search. Built on the same agentic commerce framework, it enables businesses to train agents on brand voice and data, surface Direct Offers, and support purchases through agentic checkout. In parallel, Google has also expanded Google Merchant Center data attributes to include elements such as common product questions, compatible accessories, and substitute products.

A Move to Monetize AI: Will One Protocol Prevail?

With the launch of ACP, OpenAI and Stripe are making a move to capitalize on ChatGPT’s current position as a high-traffic entry point for user intent. By enabling direct transactions inside the assistant and reducing friction in the path from intent to purchase, it positions ChatGPT not just as a discovery surface, but as a transactional layer, where purchases completed through Instant Checkout incur a small fee.

UCP, by contrast, might signal a broader strategic ambition from Google to standardize a larger portion of the commerce lifecycle and tie it closely to its existing ecosystem across Search, payments, identity, and merchant infrastructure. The goal appears to be ensuring Google remains structurally central as AI-assisted shopping begins to compete with traditional search-driven traffic.

It remains unclear whether ACP and UCP will coexist long term, converge, or compete directly for dominance. What is clear is that whichever protocol sees faster and wider adoption has the opportunity to influence the emerging “shopping layer” of the internet, including which payment providers, advertising models, and demand-generation mechanisms sit closest to agent-driven commerce.

• OpenAI enters this race with a distribution advantage, as ChatGPT is currently better positioned as a primary interface for user intent compared to Gemini.
• Google, however, has a clearer path to monetization, with mature shopping feeds, payments, wallets, and advertising infrastructure already deeply embedded in its ecosystem.

Rather than focusing on which protocol may win, the more important signal for merchants is structural: AI tools are rapidly becoming both distribution and conversion layers. As agentic commerce matures, optimization will increasingly be about how well a store can be understood, trusted, and acted upon by machines, not just discovered by humans.

What This Means for Your D2C Brand and How to Prepare

At this stage, the priority is not choosing between UCP and ACP. A multi-protocol reality is already emerging, with players like Walmart, Shopify, and Stripe supporting or endorsing both approaches. Which protocol becomes more relevant to your brand will largely depend on your existing commerce platform, payment stack, and operating regions.

More importantly, AI agents should be treated as a new commerce channel, not as a feature tied to a single vendor. Protocols, partnerships, and interfaces will evolve quickly. The underlying requirement that remains stable is the need for structured, machine-legible commerce infrastructure. Here are some tips to prepare for what is coming:

1. Establish measurement for agentic traffic

If you aren't already, begin defining how you will detect and measure AI- or agent-driven traffic and transactions. Even basic visibility helps assess whether your category or brand is seeing early agentic demand, and how urgent deeper investment should be.

2. Audit and strengthen structured product data

Review your product feeds to ensure they are comprehensive, accurate, and consistently maintained. This has long been critical for SEO and Merchant Center, but in an agentic commerce context it becomes foundational. High-quality structured data determines whether your products can be surfaced, compared, and acted upon by AI systems when protocol support becomes available for your stack or region.

3. Track your platform’s agentic roadmap

Closely monitor how major platforms are positioning themselves. Shopify, Adobe Commerce, and marketplaces like Amazon and Walmart are all approaching agentic commerce differently, and their timelines and integrations will directly affect what is feasible for your store without heavy customization.

4. Validate agent accessibility across key journeys

Test how accessible your storefront and APIs are to non-human actors. This includes crawlability, semantic HTML, clear entity relationships, and the ability for agents to traverse product, cart, and checkout flows without breaking. What works for browsers does not automatically work for agents.

5. Reduce ambiguity in policies and product relationships

Ensure policies around pricing, warranties, delivery, and returns are explicit and easy to interpret. Similarly, clearly defined relationships between products, accessories, and substitutes help agents make confident decisions rather than stopping at ambiguity.

6. Participate in pilots when available

Apply to pilot programs as they open. Early access forces learning, exposes operational gaps, and often provides preferential positioning as agent-driven commerce capabilities expand.

Optimizing Your Commerce Stack for What Comes Next

As AI agents become meaningful participants in commerce, the brands that benefit first will be those with clean data, predictable systems, and flexible platform architectures. At TMO, we work with brands on the technical foundations that make this possible, from structured product data and checkout logic to custom integrations and long-term platform optimization, helping teams build eCommerce systems that are resilient to new channels and changing buying behaviors.

If you’re reviewing the technical health of your Shopify or Magento store, planning platform improvements, or preparing for new commerce channels, we can help you assess where you stand and define a practical optimization roadmap. Reach out to us for a 1:1 discussion.