Google Bets on AI-Powered Agentic Shopping at I/O 2026
At Google I/O 2026, Google unveiled what may be its most ambitious move into e-commerce to date: a Universal Cart powered by Gemini that works across different retailers and multiple Google surfaces. The announcement signals a clear strategic direction — Google wants to become the central layer through which consumers research, select, and purchase products online, regardless of which retailer is selling them.
For shop operators, e-commerce managers, and platform developers, this development is not a distant future scenario. It is a structural shift in how the online purchase journey may be organised — and it carries direct implications for how merchants manage their digital presence, product data, and customer relationships.
What Was Announced: The Core Features
The Universal Cart allows shoppers to add products to a single, persistent cart as they browse Google Search and interact with Gemini. According to Vidhya Srinivasan, Vice President and General Manager of Ads and Commerce at Google, the cart is designed to address a practical problem: consumers shop across multiple devices, multiple sessions, and multiple retailers. The Universal Cart aggregates all of this into one place, accessible via a cart icon displayed next to a user's profile picture across Google properties.
Key features of the Universal Cart as announced include:
- Cross-retailer functionality, with named partners including Sephora, Target, Wayfair, and Walmart
- Price-drop alerts, price history tracking, and in-stock notifications
- AI-driven compatibility warnings — for example, flagging incompatible components when a user is assembling a PC
- Integration with retailer loyalty programmes and payment methods via Google Pay
- The ability to transfer cart contents to a retailer's own website for checkout if the shopper prefers
- Planned expansion to YouTube and Gmail as additional surfaces for adding products
Google also announced that the Universal Cart integrates with Gemini Spark, described as a new "24/7 personal AI agent." Through this, users can configure specific purchase parameters — preferred brands, target items, budget ceilings — and the AI agent will execute purchases autonomously when the criteria are met. These transactions run on a technology called Agent Payments Protocol (AP2), which functions as a digital approval and audit trail for AI-executed purchases.
The Infrastructure Behind It: Universal Commerce Protocol
The Universal Cart does not exist in isolation. In January 2026, Google announced the Universal Commerce Protocol (UCP), an open standard co-developed with major retailers including Walmart, Shopify, and Target. UCP is designed to standardise the entire AI shopping journey — from product research and cart management through to payment and post-purchase customer service. By April 2026, Amazon, Meta, Microsoft, Salesforce, and Stripe had joined the committee governing UCP.
This level of industry participation suggests that UCP is gaining traction as a foundational standard for agentic commerce. OpenAI has a competing version, indicating that the infrastructure layer of AI-driven shopping is itself becoming a competitive battleground.
The "DoorDash Problem": Why Merchants Should Pay Attention
The commercial logic of Google's approach raises a significant tension for retailers. If consumers can research, compare, and purchase products entirely within Google's ecosystem, there is reduced incentive for them to visit a merchant's own website at all. This dynamic — referred to in the source article as the "DoorDash problem" — mirrors what happened in food delivery, where aggregator platforms captured the customer relationship while restaurants supplied the product.
Srinivasan characterises Google's role as that of a "matchmaker" rather than a merchant of record, and states that Google currently does not take a commission or cut of sales generated through the Universal Cart. She also notes that post-purchase issues would be handled between the customer and the retailer directly, not through Google. Whether this model holds as the platform scales remains an open question.
The concern is not hypothetical. Amazon took legal action against AI company Perplexity for enabling purchases through its Comet AI browser, and brands are already adjusting their content strategies to improve visibility in AI-powered search results and chatbot recommendations.
Practical Implications for Shopware Merchants and E-Commerce Operators
While Shopware is not mentioned in Google's announcement, the trends it describes have direct relevance for any merchant operating an online shop. Based on the developments described in the source material, the following areas merit attention:
- Product data quality: AI agents rely on structured, accurate product information to make decisions and flag compatibility issues. Clean, complete product feeds become more critical as agentic shopping scales.
- AI search visibility: As the article notes, merchants are already adapting their online presence to improve their chances of being surfaced by AI chatbots and agents. This is becoming a meaningful component of digital marketing strategy.
- Platform integration readiness: The UCP standard is gaining broad industry backing. Merchants and developers should monitor whether Shopware integrations or plugins emerge that support UCP-compatible product feeds or cart handoffs.
- Loyalty and payment data: Google's integration of loyalty programmes and payment methods into the Universal Cart means retailers who participate may gain access to better-informed shoppers, but at the cost of reduced direct interaction at the checkout stage.
Outlook: Trust, Adoption, and Unanswered Questions
Google's vision of agentic shopping raises questions that the company acknowledges are still being worked through. What happens when an AI agent purchases an item within a user's stated price limit, but total costs including tax and shipping exceed a comparable alternative? Will retailers introduce specific policies for AI-executed purchases? And perhaps most fundamentally: will consumers trust an AI agent to spend their money autonomously?
Srinivasan's framing — that merchants benefit in the long run only if consumers also benefit — suggests Google is aware that retailer participation depends on demonstrating genuine value, not just volume. The consumer trust challenge, however, remains substantial. Outsourcing purchasing decisions to an AI agent represents a significant behavioural shift, and adoption will depend on whether the technology proves reliable and transparent in practice.
For e-commerce professionals, the key takeaway from Google I/O 2026 is not that agentic shopping has arrived in full force, but that the infrastructure supporting it is being built at scale — and the time to understand its implications is now, not when it reaches mainstream adoption.