Follow ZDNET: Add us as a preferred source on Google.
ZDNET’s key takeaways
- Google’s Universal Cart lets shoppers purchase products from multiple retailers in one place.
- Gemini’s agentic AI runs in the background to suggest purchases.
- The goal is to automate purchases, streamline checkout, and predict consumer behavior.
At Google’s I/O developer conference on Tuesday, the company introduced a slew of updates to search, including a new feature called “Universal Cart”, an AI-powered shopping assistant that consolidates your shopping into one place under Google’s Universal Commerce Protocol. One cart, multiple retailers.
The UCP is an open standard for commerce and agentic AI, co-developed alongside major retailers such as Target, Shopify, Wayfair, and Etsy. It allows them to operate on Google Pay while still giving customers access to retailer-specific data, such as loyalty programs or credit cards.
Also: Everything we saw at Google I/O: Gemini 3.5, Android XR glasses, Spark, and more
Universal Cart gives Google’s AI access to your product selections from all over its ecosystem: YouTube, Gmail, Gemini, or search, allowing it to provide insights on what you’re buying, make suggestions, and open the door for all sorts of other interactions.
In a preview call ahead of I/O, Vidhya Srinivasan, Google’s VP of Ads and Commerce, said these features will “make shopping more fun.” What she likely means is that it reduces the barriers between “Add to cart” and “Checkout”. Retailers want this to be as frictionless as possible — instant, even — and as personalized as possible.
The agentic AI can certainly be useful. In a live demo during I/O, Srinivasan showed a shopper adding a CPU and motherboard to their cart, only to be notified by the AI that the two devices aren’t actually compatible. Good call. In another, it prompted the user to take advantage of a discount by using a different credit card.
All of this is intended to be automatic. If you’re shopping on Google and add something to the cart, it works in the background, searching for better deals, letting you know if a sale price is actually worth it, or highlighting any specific sale information — all helpful actions that aid the consumer. But it’s also tracking your behavior, keeping tabs on what you’re looking at, and predicting future purchases.
Also: Google’s new AI Search box is here – along with agents and 5 more upgrades
Google’s Universal Cart is just the tip of the iceberg. I saw Google demo a similar feature back in January during a press demo of its Auto Browse feature in Chrome, where the user gives the browser permission to take action on their behalf. In the demo, the user showed Gemini a photo of some party decorations, instructing it to find the products on the web.
After analyzing the image, it located the streamers, balloons, and decorations and added them to the cart. In theory, with Google’s UCP, you wouldn’t have to checkout in separate tabs on Etsy, Amazon, and Walmart — it would all be in one place. Easy.
All of these features work together to make digital shopping more seamless, because it needs to be if retailers want to increase conversions. Google continues to emphasize how Gemini and agentic AI excels at carrying out “digital laundry” — that is, handling routine tasks — and these features handle those shopping tasks nicely.
Also: Generative AI was everywhere at Google I/O 2026, but who is it for? – Video
The features can be very helpful, especially if you know how to leverage the agent effectively or if they alert you to specific inquiries. If you buy the same kind of toilet paper every month, Gemini is there to make sure you don’t forget, and will add it to your cart — and make the purchase — itself.
The ultimate goal is speaking to the AI in natural language and giving it permission to carry out actions on your behalf, and automating routine purchases is the first step. What could be more fun than AI agents spending your money while you sleep?
