Our Blog

Blog Index 

Google Tests Buying Directly From Flipkart via Gemini and AI Mode in India: What Agentic Shopping Means for Indian Sellers

Posted on 27th Sep 2026 12:04:29 in Artificial Intelligence, Machine Learning

Tagged as: Google, Flipkart, Gemini, AI Mode, agentic commerce, India, e-commerce, AI shopping

Google has started testing something it has been building toward all year: letting shoppers in India complete a purchase without leaving its AI interfaces. According to TechCrunch, select Flipkart product listings inside Gemini and Google's AI Mode in Search now carry a "Buy" button that opens a Flipkart-branded checkout flow directly inside the AI experience — no redirects, no switching between apps, no bouncing out to a browser tab to finish a payment.

The test is deliberately small: a limited set of users and a narrow catalogue covering smartphones, electronics, and mobile accessories. But the timing is not accidental. Google plans to widen the rollout in October, right as India's festive shopping season — the stretch when Flipkart and Amazon fight their biggest sales battles of the year — swings into full gear.

It is the clearest signal yet that the competition between Google, OpenAI, Meta, and Amazon is shifting from which AI gives the best product recommendations to which AI actually completes the purchase. For Indian sellers and small businesses, that shift matters more than it might first appear.

What Google Is Testing, and How It Works

The mechanics, as reported by TechCrunch: when a shopper asks Gemini or AI Mode a shopping question, some Flipkart listings now render a "Buy" button alongside the product card. Tapping it pulls up a checkout flow carrying Flipkart's branding, still within the AI interface. The test is limited to a subset of users and a small selection of products — mostly smartphones, electronics, and mobile accessories — while other users continue to see Flipkart listings without the buy option.

Asked about the test, a Google spokesperson said the company is "always testing new features and experiences to help people discover and connect with businesses more easily," adding that Google runs experiments regularly and had no further details to share. Flipkart did not immediately respond to a request for comment.

One detail worth noting: this checkout is not the Google-hosted experience the company demonstrated earlier. Instead of Google running the payment inside Gemini, users see a Flipkart-branded checkout flow. It is not yet clear what technology powers the test, but the structural difference suggests Google is experimenting with letting large retailers keep control of the transaction while it still owns the surface where the purchase begins. Notably, Amazon listings appear alongside Flipkart products in the same AI results but do not offer the buy option — a reminder that platforms are still negotiating who gets to sit where in the AI shopping funnel.

The Bigger Play: Universal Commerce Protocol and the Agentic Shopping Race

The Flipkart test is the consumer-facing tip of a much larger infrastructure project. Earlier this year, Google introduced the Universal Commerce Protocol (UCP), an open standard designed to let AI agents interact with retailers across the entire shopping journey — discovery, checkout, and post-purchase support — without every retailer needing separate integrations for every AI assistant. UCP was developed with input from companies including American Express, Shopify, Etsy, Target, Wayfair, Walmart, and Flipkart, and it has received backing from more than 20 firms across payments, retail, and e-commerce platforms.

Google's first UCP checkout implementations launched in the United States, on eligible product listings in AI Mode and Gemini, with payments handled through Google Pay using details stored in Google Wallet, and PayPal support planned later. In that model, the retailer remains the seller of record. The Flipkart test in India appears to be a different flavour — retailer-branded checkout rather than a Google-hosted one — which may make it easier for large marketplaces with their own payment stacks to participate.

Google is not moving alone. OpenAI and other AI companies are racing to add commerce capabilities to their assistants, trying to convert shopping conversations into transactions. The friction is already showing: Amazon recently blocked Meta's personal AI agent, Muse, from accessing its shopping site, citing privacy and security risks — a sign that the world's largest retailers are wary of becoming interchangeable inventory for someone else's AI.

Flipkart, meanwhile, is building from its own side. Its fintech arm Super.money has begun rolling out AI agents that can autonomously shop for products and buy gold when prices dip to levels customers set in advance, and its roughly 20 million monthly active users are the testing ground. Super.money CEO Prakash Sikaria told Bloomberg (via CNBC-TV18) that the company plans to cover its entire customer base within two months and expects agentic use cases to contribute 30 to 40 percent of revenue within three to four years. "Because we are a very value-conscious country, we believe that price-led or value-led agents will do well," Sikaria said. Google's own YouTube is pushing in the same direction: at its Made on YouTube event, it expanded shopping on the platform, including the growth of its YouTube Shopping affiliate programme in India with Amazon and new partners such as AJIO, Meesho, Snapdeal, and Tata CLiQ.

Why India Comes First

India is the world's second-largest internet market, with more than a billion internet subscribers, and its festive season is the single biggest demand event on the e-commerce calendar. It is also a market where Google is not a stranger at the transaction layer: the company invested about $350 million in Flipkart in 2024, taking a minority stake alongside its technology partnership with the Walmart-owned marketplace.

India also happens to be where the economics of agentic shopping are most interesting to test. Indian shoppers are famously value-driven, quick to compare prices, and heavily reliant on cash on delivery — a recent Indian Express report noted that while 87 percent of shoppers find AI-assisted shopping faster, 88 percent still choose cash on delivery. An AI agent that reliably finds the best price and reports it inside the chat fits how India shops; an AI checkout that cannot route a cash-on-delivery order fits less well. How Google and Flipkart handle that gap will decide how quickly the Buy button moves from experiment to habit.

What It Means for Indian Sellers and Small Businesses

For sellers, the strategic question is simple: if a growing share of purchases starts inside AI assistants rather than on marketplace pages or search results, who gets shown — and on what basis?

First, product data becomes the storefront. AI systems can only recommend and, eventually, transact on products whose titles, specifications, pricing, availability, and return policies are structured and machine-readable. Sloppy listings and inconsistent catalogues risk being invisible to the agent economy the same way a site without search-friendly content is invisible on Google today.

Second, trust signals travel with you. An AI that has learned to buy on a customer's behalf will prefer sellers and platforms with strong ratings, reliable delivery records, and frictionless returns, because those are the parameters that make autonomous purchasing safe to attempt at all.

Third, checkout is being rebuilt. From UPI to agent wallets to retailer-branded AI checkouts, the payment layer underneath commerce is changing fast. Small businesses that keep their payment options modern — and that check whether their catalogues are readable by the tools their customers actually use — will be positioned to ride the wave rather than chase it.

And fourth: this is still early. The current test covers a handful of product categories and a fraction of users. The October expansion, however, lands squarely in the festive season, which means millions of Indian shoppers will get their first taste of buying through an AI assistant exactly when they are most willing to spend.

What to Watch Next

Over the coming weeks, three things will show how fast agentic shopping is really moving in India: whether Google expands the Buy button beyond smartphones and electronics into higher-consideration categories; whether other large Indian retailers beyond Flipkart get the same treatment; and whether Google's UCP-based Google-hosted checkout and retailer-branded flows like Flipkart's converge into a single standard. For now, the takeaway for Indian businesses is straightforward: the AI assistant is no longer just the place where shopping starts. It is rapidly becoming the place where shopping happens.

Sources

Recent Posts

Catgories

Archives

whatsapp me