The Verge said in an X post that Google’s AI shopping tool will come first to Google Search and the Gemini app. Google’s own announcement, published on November 13, 2025, provides more detail: conversational shopping is coming to AI Mode in Search and Gemini, with product recommendations, comparison tables, prices, reviews and inventory information.
The placement is strategically important. Search already captures the moment when people express what they want, while Gemini is designed for longer, conversational requests. By putting shopping in both products, Google can treat product research as an extension of a user’s existing interaction rather than asking consumers to visit a separate shopping destination.
That could give Google control over more of the purchase journey. A shopper might describe a budget, a preferred brand and a set of requirements in ordinary language. The system could then narrow the field, explain trade-offs and display current information in a format that reduces the need to open multiple retailer pages. The value for Google would not be limited to answering questions. It would come from influencing which products users consider and where they ultimately buy them.
Search becomes a commercial interface
According to Google’s announcement, the new experience will combine recommendations with comparison tables, prices, reviews and inventory data. Those features position the tool as more than a product search box. They create an AI-mediated layer between a shopper and the retail market.
Google also described AI store calls, which can gather information from stores, and agentic checkout for eligible merchants in the United States. The announcement therefore points toward two different business opportunities. The first is assistance, where Google helps users find and evaluate products. The second is transaction support, where the company takes a more active role in helping complete a purchase.
The distinction matters for competition. Product discovery keeps Google close to its traditional search business, where visibility and commercial intent are central assets. Agentic checkout would move the company closer to the transaction itself. That could increase the strategic value of Google’s shopping data, while also making the quality of its recommendations more consequential.
For retailers, the opportunity is access to customers who are ready to buy without forcing them through conventional search and category pages. The risk is that the retailer relationship becomes less visible. If Gemini or AI Mode summarizes products and presents a short list, merchants may have less control over how their catalogs, pricing and brand differences are interpreted.
Google’s announcement does not establish the full commercial terms of the rollout. It does not specify a launch date, the complete list of supported markets, the breadth of retailer coverage or whether agentic checkout will be available beyond eligible U.S. merchants. Those details will help determine whether the product is a broad retail infrastructure play or a limited shopping feature.
Trust is the central execution risk
The move also places pressure on Google to explain how its commercial incentives will operate inside AI-generated recommendations. A conventional search page can distinguish advertising, organic results and merchant listings through visible formats. A conversational answer may compress those distinctions into a recommendation that feels personal and authoritative.
That creates questions about sponsored placement, product data, reviews and affiliate relationships. It also creates operational risks. A system can misunderstand a shopper’s budget, rely on outdated inventory or recommend a product that does not match the user’s requirements. In an ordinary search session, the user may correct the mistake by reviewing several links. In an agentic shopping flow, the recommendation could carry more weight because the assistant appears to have already done the comparison.
U.S. Senator Elizabeth Warren has raised those concerns. In a letter to Google, Warren criticized the company’s AI shopping partnerships and warned that sensitive user data could be used to manipulate purchases, personalize prices or promote upsells. She asked Google to explain its safeguards and consent practices.
Warren’s position is the only political response identified in the supplied material, and it is critical. No opposing public view was found in those sources. That does not resolve the underlying questions, but it does make the company’s disclosures particularly important as the product moves closer to transactions.
Independent research cited in the available material also points to a reliability gap. In the ShoppingComp paper, researchers Huaixiao Tou, Ying Zeng and colleagues reported that current shopping agents performed poorly on realistic product-retrieval, report-generation and safety-critical tasks. The researchers argued that major reliability problems remain before such systems can be deployed at scale.
Google’s advantage depends on execution
Google starts with assets that many AI competitors would struggle to match in combination: a dominant search interface, the Gemini app, shopping-related product information and relationships across the web. Those assets give the company distribution and a potentially large stream of commercial intent.
Distribution alone, however, will not create a durable shopping moat. Google must make recommendations accurate, explainable and current while preserving enough value for merchants to keep their product data accessible. It must also show users that conversational convenience does not mean hidden commercial influence.
For retailers, Google’s system could become an important new channel for discovery. For shoppers, it could reduce the time required to compare products. For competing AI companies, it raises the standard for assistants that want to participate in commercial decisions rather than simply answer questions.
The strategic test is whether Google can connect conversational intelligence to trustworthy retail data without losing user confidence. Search and Gemini give it the strongest possible starting points. The next advantage will come from execution: dependable recommendations, transparent incentives and a checkout experience that users believe is working for them rather than merely selling to them.
Rebeca Smith is not a person. No notebook, no deadlines, no face behind the name — just a byline this newsroom publishes under. Here is the production line underneath it, because a name beside a portrait reads like a journalist, and this one is not one.
The models. Writing: gpt-5.6-luna and qwen3-max. Out on the live web: gpt-5.6-luna and gpt-5.6-terra. Pictures: gpt-image-1 and gpt-image-1-mini. Swap one in the newsroom and this line swaps with it — it is read off the machines, not typed here.
How a story is made
- Research. The searching model reads around the story, pointed at primary sources — the filing, the post, the repository — rather than at somebody else's write-up of them.
- Writing. The writing model drafts it against what was found, at Rebeca Smith's usual length and in Rebeca Smith's usual register.
- The loop. A reviewer reads the draft and sends it back with notes. Then reads it again. A piece can go round several times before it leaves the building.
- Enrichment. A quotation has to appear word for word on the page it is taken from. A chart may only use figures that appear in the source it cites. Whatever fails is dropped, and the reason is kept.
- Fact check. A last pass hunts for claims the article makes and its sources do not.
- A human stop. Sensitive subjects are held for a person to read before publication, and a person can kill any of it at any point.
If that sounds less like a newsroom and more like a factory: quite. It is called Press Factory.
This article was generated using AI and published automatically without human pre-publication review.
How this article was made
The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.