January 12, 2026

Google just ended the protocol wars before most people realized they had started. At NRF 2026 yesterday, Sundar Pichai unveiled the Universal Commerce Protocol (UCP), an open-source standard co-developed with Shopify, Etsy, Wayfair, Target, and Walmart. More than twenty other companies have endorsed the standard, including Adyen, American Express, Best Buy, Flipkart, Home Depot, Macy's, Mastercard, Stripe, Visa, and Zalando. The protocol establishes a common language for AI agents to discover products, execute checkouts, apply discounts, and track orders across any participating retailer. Pichai framed the stakes bluntly: "The industry needs a protocol that works at global scale and takes into account the nuances of commerce journeys." OpenAI's competing Agentic Commerce Protocol, developed with Stripe, remains open-source and technically viable. But Google just recruited most of the retail industry to its side.

The technical architecture matters. UCP supports three integration methods, including traditional REST APIs, Google's Agent-to-Agent (A2A) protocol, and Anthropic's Model Context Protocol (MCP). Businesses publish capabilities via a standardized /.well-known/ucp JSON manifest, allowing agents to dynamically discover what a merchant can do without hardcoded integrations. The protocol separates payment instruments from handlers, enabling interoperability across Google Pay, Shopify Pay, and any processor supporting the Agent Payments Protocol (AP2) for cryptographic proof of user consent. This isn't a walled garden. It's infrastructure designed to let any compliant agent transact with any compliant merchant. Google says UCP will power checkout directly in AI Mode in Search and the Gemini app, letting users buy without leaving the conversation.

The payment networks are moving even faster to close the last-mile gap. Fiserv announced it will be one of the first major processors to integrate Mastercard's Agent Pay Acceptance Framework at scale, enabling AI-initiated purchases to be authenticated, tokenized, and settled through existing card rails. The framework includes mechanisms to distinguish authorized AI agents from malicious automation. It evaluates whether an agent has permission to act on a user's behalf and whether the transaction falls within predefined parameters. Fiserv has simultaneously partnered with Visa to support the Trusted Agent Protocol, signaling that multi-network support for agentic commerce is becoming table stakes. Visa predicts that millions of consumers will use AI agents to complete purchases by the 2026 holiday season.

Microsoft entered the checkout race with its own announcement. Copilot Checkout lets shoppers complete purchases directly inside the AI chatbot with no redirect and no friction. The company is betting its enterprise relationships can differentiate against OpenAI, Google, and Amazon. Microsoft also launched Brand Agents, pitched as a turnkey solution for Shopify merchants to add AI assistants to their storefronts, plus new Dynamics 365 Commerce MCP Server capabilities that expose catalog, pricing, inventory, carts, and fulfillment data as MCP-enabled endpoints. If your existing commerce stack runs on Microsoft, agentic capabilities are now native.

Amazon is playing a more complicated game. CEO Andy Jassy disclosed that Rufus, Amazon's AI shopping assistant, is on pace to generate $10 billion in incremental annual sales. Some 250 million shoppers have used it, with interactions up 210% year-over-year. Customers who engage with Rufus are 60% more likely to complete a purchase. But Jassy also told analysts that most third-party AI shopping agents fail to provide a satisfactory customer experience, citing poor personalization and inaccurate pricing and delivery estimates. Amazon's proprietary agent works. External ones don't. Yet Jassy acknowledged Amazon is "having conversations with and expect, over time, to partner with third-party agents". Even the company most invested in keeping commerce within its walls sees the protocol layer becoming unavoidable.

The adoption curve is steepening across the industry. NVIDIA's third annual State of AI in Retail and CPG survey found that 47% of respondents are using or assessing agentic AI, with 20% reporting agents already active in production and another 21% deploying within the year. Internal workflow automation and knowledge retrieval lead at 59% adoption, followed by customer support, employee assistants, and personalized marketing at 48% each. Speed and efficiency gains rank as the primary goal for 57% of deployments. NVIDIA's analysis suggests the "truly disruptive impact of agentic AI will hit retail supply chains and operations first." Autonomous agents handling real-time inventory rebalancing, dynamic pricing, and vendor negotiations offer the most measurable ROI. The talent constraint is real. The AI talent shortage intensified from 31% to 46% year-over-year and is now the primary implementation barrier.

The strategic picture is crystallizing. Constellation Research's NRF coverage frames it cleanly: retailers are trying to figure out agentic AI-driven commerce while keeping frontline workers engaged enough to deliver customer experience. Those are the two defining themes of this year's conference. Pichai's ambition is explicit: "A future where you are able to build strong relationships with your customers that long outlast any single search or purchase." What he didn't say is that in that future, Google sits between the customer and the merchant. The retailers who endorsed UCP know this. They signed anyway, because the alternative is building N×N integrations with every agent provider while OpenAI, Microsoft, and Amazon race to capture intent. If 2025 was the year companies laid groundwork for agentic commerce, 2026 is the year they're choosing which standard to build on. The window for hedging is closing.