Agentic commerce is an emerging model of online buying and selling in which AI agents research, negotiate, and complete purchases on behalf of consumers or businesses — often without direct human intervention at each step. The human sets intent, approves decisions, and retains control. The agent handles execution.
This is not a theoretical concept. As of March 2026, Google has launched the Universal Commerce Protocol (UCP), an open standard that enables AI agents to interact with merchants across discovery, checkout, and post-purchase. OpenAI launched Instant Checkout in ChatGPT via its Agentic Commerce Protocol built with Stripe. Visa introduced its Trusted Agent Protocol, and Mastercard deployed Agent Pay.
How Agentic Commerce Works
In traditional ecommerce, a human browses a website, adds items to a cart, enters payment details, and clicks “Buy.” In agentic commerce, the workflow looks fundamentally different:
- The human declares intent. Example: “Find me a camping tent under $150 and have it delivered by Friday.”
- The AI agent interprets and acts. It accesses structured product data via protocols like UCP, filters by price, specs, and delivery availability, then compares options across merchants.
- The agent presents recommendations or completes the purchase — depending on the user’s permission level and the transaction type.
The agent doesn’t browse a website the way a human does. It calls APIs. It reads structured data. It operates through machine-to-machine protocols, not visual interfaces.
The Protocols Making It Real
Three foundational protocols are now shipping that enable this architecture:
- UCP (Universal Commerce Protocol) — Launched by Google at NRF 2026 and endorsed by Visa, Mastercard, Shopify, Stripe, Best Buy, Target, Walmart, and others. UCP provides a common language for agents and systems to operate together across the shopping journey. In March 2026, Google added cart support and product catalog access to UCP. Currently available to eligible U.S.-based merchants, with global expansion planned throughout 2026.
- A2A (Agent-to-Agent Protocol) — Originally developed by Google, now under the Linux Foundation. Enables agents from different organizations to communicate and collaborate.
- MCP (Model Context Protocol) — Developed by Anthropic. Provides a standard way for AI models to access tools and data sources.
Who Is Already Using Agentic Commerce
According to a 2026 IBM Institute for Business Value study, 45% of consumers already use AI for some part of the buying journey. Major deployments include:
- Google AI Mode — UCP-powered checkout embedded in Google Search for qualifying merchants.
- ChatGPT Shopping — OpenAI’s product discovery and checkout experience, currently live with Etsy and expanding to Shopify merchants including Glossier, SKIMS, and Vuori.
- Shopify Agentic Storefronts — Shopify merchants can now expose their catalog and checkout flow to AI agents via ChatGPT.
The payment networks are moving in parallel. Visa’s Trusted Agent Protocol and Mastercard’s Agent Pay both use public key infrastructure (PKI) to enable secure agent-driven transactions, according to Payments Dive.
What Makes It Different From Regular AI in Ecommerce
AI has been used in ecommerce for years — recommendation engines, chatbots, dynamic pricing. Agentic commerce is different because the AI agent is the transacting party. It doesn’t just recommend a product; it negotiates, selects, pays, and handles post-purchase on the consumer’s behalf.
As McKinsey described it: agentic commerce moves us toward a world in which AI anticipates consumer needs, navigates shopping options, negotiates deals, and executes transactions — all in alignment with human intent yet acting independently via multistep chains of actions enabled by reasoning models.
Why This Matters for Merchants
If your product data isn’t structured for machine consumption, AI agents won’t find you. If your checkout doesn’t support agent protocols, AI agents can’t buy from you. The shift from human-browsable storefronts to machine-readable commerce surfaces is the most significant architectural change in online retail since the introduction of mobile commerce.
Merchants preparing for this shift should focus on: structured product data (UCP-compatible schemas), agent-accessible checkout flows, and machine-readable policies (shipping, returns, warranties) that agents can parse without human interpretation.
Related Reading
- Agentic Commerce vs Traditional Ecommerce: What Actually Changed
- What Is Agentic AI Checkout? How AI Completes Purchases
- Visa vs Mastercard: The Agentic Commerce Compliance Race
- How AI Shopping Agents Actually Work: A Technical Walkthrough
- Cookieless Affiliate Tracking: How Attribution Works When AI Agents Shop
- Verifiable Intent: The Privacy Architecture Securing AI Commerce
- AI Agent Verification: How Merchants Know They’re Dealing With a Real Agent
- Where Can You Use Agentic Commerce Today? A 2026 Reality Check
- Is Agentic Commerce the Same as Agentic AI?
Leave a Reply