What Are ACP, UCP, and MCP? The Agentic Commerce Protocols Explained

ACP (Agentic Commerce Protocol), UCP (Universal Commerce Protocol), and MCP (Model Context Protocol) are the three protocols powering AI-driven commerce. Each operates at a different layer of the same transaction stack, and they are designed to coexist rather than compete.

What Is MCP?

MCP is the foundation layer. Created by Anthropic and now governed by the Linux Foundation, it provides a general-purpose protocol for connecting AI agents to external tools and data sources including commerce APIs. It is not commerce-specific. Think of MCP as the universal adapter that lets an AI agent query a product catalogue, check inventory, pull order history, or trigger any backend action. Enterprise AI tools from Salesforce, Adobe, and SAP use MCP for catalogue integration.

What Is ACP?

ACP is the checkout and payment layer, built by Stripe and OpenAI. It handles commerce flows inside ChatGPT and went live in production in late 2025 with PayPal and Worldpay as payment partners. ACP provides a REST-based Checkout API for cart management and a Delegate Payment API for secure payment tokenisation through PSPs like Stripe. The minimal surface area is intentional: ACP focuses only on the moment the cart is ready and the user wants to pay.

What Is UCP?

UCP is the broadest of the three. Launched by Google with Shopify and Walmart at NRF 2026, it covers the full commerce journey from product discovery through post-purchase. UCP standardises this through a single abstraction layer, collapsing the N-times-N integration complexity that emerges when every retailer must build separate connections for every AI surface. Google describes UCP as compatible with existing protocols including ACP.

How They Work Together

A single AI-driven purchase may use all three. MCP handles tool connectivity and authentication. UCP handles product discovery and comparison. ACP handles checkout and payment tokenisation. The channel dictates which protocol matters most: ChatGPT uses ACP, Google AI Mode and Gemini use UCP, and enterprise tools rely on MCP for catalogue access. Most serious retailers will end up supporting at least two protocols.

SEO and Commerce Implications

These protocols shift where visibility and conversion happen. When an AI agent handles discovery and checkout, the traditional search-to-click-to-PDP funnel is compressed. Merchants must expose structured product data through protocol-compatible endpoints: Product schema with price, availability, and GTIN at minimum, with MCP servers or Shopify UCP feeds for serious agentic commerce volume. Agents that cannot retrieve accurate product data reliably will surface competitors who can.