The Battle Nobody Saw Coming
For years, the agentic commerce conversation was about *whether* AI would start buying things for consumers. That question is settled. OpenAI's ChatGPT processes 50 million shopping queries per day. Amazon's Alexa for Shopping replaced Rufus on May 13, 2026, inheriting 300 million users overnight. Google's AI Mode enables autonomous checkout through Google Pay.
The new question is far more consequential: when an AI agent buys something for you, whose rails does the transaction run on?
This is the Protocol Wars. Three of the most powerful technology companies in the world have each launched competing standards for how AI agents discover products, negotiate purchases, and complete checkout. The winner will not just own a feature. They will own the infrastructure layer of the next era of commerce, the same way Visa and Mastercard own the infrastructure layer of payments today.
Three Protocols, Three Philosophies
Google: The Universal Commerce Protocol
Google announced UCP at NRF in January 2026, co-developed with Shopify and endorsed by over 20 major commerce players including Walmart, Target, Etsy, Wayfair, Best Buy, Flipkart, Mastercard, Visa, Stripe, Adyen, American Express, and Zalando.
UCP's philosophy is genuinely open. Any AI agent, whether built by Google or a competitor, can use the protocol to discover products, check availability, and initiate checkout across any participating merchant. Google's AI Mode enables agentic checkout: Google buys items on the merchant's site using Google Pay on the user's behalf.
Google also launched "Business Agent," a virtual sales associate that lets shoppers chat with brands directly on Search, creating an AI-intermediated conversation layer between consumer intent and merchant response.
The strategic logic: Google does not need to own the checkout. It needs AI agents to keep using Search as their discovery layer. If UCP becomes the standard, every AI shopping query, regardless of which agent initiates it, routes through Google's commerce graph.
OpenAI: The Agentic Commerce Protocol
OpenAI launched "Buy it in ChatGPT" on February 16, 2026, initially with an ambitious Instant Checkout feature. The underlying standard, the Agentic Commerce Protocol, was co-developed with Stripe.
By March, OpenAI pivoted. Instant Checkout struggled with merchant onboarding, data accuracy, and multi-item carts. The revised approach lets merchants own their checkout experience while AI handles discovery. Partners include Target, Sephora, Nordstrom, Best Buy, Home Depot, and Wayfair. Over 1 million Shopify merchants are coming online. PayPal's ACP server will bring tens of millions of additional small businesses onto the platform.
ACP's philosophy is merchant-friendly. The protocol recognizes that merchants do not want to cede checkout to an AI intermediary. They want the traffic, the data, and the customer relationship. ACP gives them that while still enabling AI discovery. ChatGPT traffic converts at 1.81% versus 1.39% for non-branded organic, with 10.3% higher revenue per session.
Amazon: The Closed Ecosystem
Amazon chose a different path entirely. On May 13, Amazon retired Rufus and launched Alexa for Shopping, merging it with Alexa+ and inheriting 300 million users. The system features account memory, personalized recommendations, price tracking, and auto-purchase at target prices.
The most aggressive feature is Buy for Me. When a product is unavailable on Amazon, the AI agent locates it on a third-party retailer's site and completes the purchase using the customer's encrypted payment information. The Shop Direct program covers 100 million products from 400,000+ merchants. Free for all Amazon users, no Prime required.
Amazon's philosophy: own everything. The consumer never leaves Amazon's ecosystem, even when buying from a competitor's inventory.
Why Protocols Matter More Than Products
The protocol that wins becomes the default interface between all AI agents and all commerce. Consider what happened with payment card networks. Visa and Mastercard do not sell products. They do not set prices. But they capture 2-3% of nearly every consumer transaction because they own the protocol layer connecting buyers, sellers, and banks.
Commerce protocols are headed for the same dynamic. The protocol AI agents default to for product discovery and checkout initiation will capture enormous value, not through transaction fees initially, but through data, attention routing, and eventually monetization.
The stakes are higher than payments. A payment protocol just moves money. A commerce protocol controls *what gets bought*. When ChatGPT recommends three products via ACP, or Google's AI Mode selects items via UCP, or Alexa auto-purchases through Buy for Me, the protocol shapes the purchase decision itself.
The Merchant's Dilemma
Brands now face a fragmented agent ecosystem:
You need to be discoverable by all three protocols simultaneously. A consumer using ChatGPT will not find you if your data is not ACP-compatible. A Google AI Mode user will not see you outside UCP's merchant graph. An Alexa user will not be directed to you outside Amazon's catalog.
Each protocol has different data requirements. ACP leans on Stripe's merchant infrastructure. UCP integrates with Google's Shopping Graph. Amazon uses its own catalog plus third-party scraping. Maintaining optimized, structured product data across all three is now a critical capability.
Attribution is fragmenting. When an AI agent completes a purchase, the attribution path is opaque. Traditional analytics cannot answer which agent drove the sale, and new measurement frameworks do not exist yet.
Pricing consistency is fully exposed. All three systems compare prices across merchants in real time. A price discrepancy between your Amazon listing, Shopify store, and Walmart listing will be surfaced instantly. Brands need real-time pricing intelligence across all channels. Platforms like BrandBaazar that monitor 50+ marketplaces become essential infrastructure in a multi-protocol world.
Who Wins?
Probably nobody wins completely.
Google wins discovery. UCP's open approach and endorsement list positions it as the discovery protocol. Google's Shopping Graph is the most comprehensive product data source. Willingness to let merchants own checkout reduces adoption friction.
OpenAI wins the high-intent consumer. ChatGPT's 50 million daily shopping queries and 1.81% conversion rate represent deeply intentional research. ACP's merchant-friendly approach means premium brands will prioritize it.
Amazon wins the habitual purchase. Alexa for Shopping with Buy for Me keeps consumers inside Amazon's ecosystem. For replenishment, commodity purchases, and Prime-habituated consumers, Amazon's closed system will dominate.
What Brands Should Do Now
Audit your structured product data across all surfaces. Is it complete, accurate, and machine-readable across Amazon, Shopify, Google Shopping, and every other channel? Missing data means missing agent recommendations.
Monitor protocol adoption by your retail partners. Which platforms have implemented UCP, ACP, or both? Your product visibility now depends on your retail partners' protocol integrations.
Invest in real-time competitive intelligence. When AI agents surface every competitor simultaneously, pricing discrepancies cost you sales in real time. Tools like BrandBaazar's marketplace intelligence give you the cross-platform visibility that protocol-fragmented commerce demands.
Build for protocol-agnostic discovery. Do not bet on one protocol winning. Build data infrastructure to serve all three: structured schema for Google, API-accessible catalogs for OpenAI, and optimized Amazon listings for Alexa.
The Protocol Wars are just beginning. But the brands that recognize what is at stake, not which AI agent consumers prefer but which infrastructure layer controls the purchase decision, will be best positioned for whatever the outcome is.