Google tests in-AI shopping checkout with Flipkart in India

Google has begun testing a new way for people in India to buy products from Walmart-owned Flipkart directly inside Gemini and Google’s AI Mode, a move that could turn AI search from a product-discovery tool into a shopping checkout channel. The experiment is limited for now, but it signals Google’s push to make its AI interfaces more commercially useful ahead of India’s busy festive shopping season.

The test adds a “Buy” button to selected Flipkart listings shown in Gemini and AI Mode, allowing users to jump straight into a Flipkart checkout flow without leaving Google’s AI experience. The rollout is currently narrow, covering only some users and a small set of items, but Google is expected to widen access later in October, according to people familiar with the matter.

The development matters because it shows Google is trying to deepen AI’s role in commerce at the same time as rivals such as OpenAI race to add shopping features of their own. Instead of merely recommending products, AI assistants are increasingly being positioned as intermediaries that can complete transactions.

What Google is testing in India

Google’s experiment centers on a direct-purchase flow for Flipkart products surfaced in Gemini and AI Mode. In practice, that means some shoppers who see eligible listings can tap a Buy button and proceed into checkout rather than being sent to a separate browser page or retailer app.

That shift is important. AI shopping tools have largely focused on discovery: answering questions, comparing products, and generating recommendations. Google’s India test suggests the company wants to go one step further by keeping users inside its AI interface as they move from browsing to buying.

How the test appears to work

In the version observed by TechCrunch and described by people familiar with the test, the checkout flow that opens after tapping Buy is branded for Flipkart, not Google. That differs from a separate Google demonstration earlier this year that showed a Google-hosted checkout experience powered by its commerce technology.

Google has not publicly explained what backend system is powering this specific trial. But the visible experience suggests the company is experimenting with a retailer-managed checkout embedded into an AI-driven product journey.

A Google spokesperson said the company is constantly trying new features and experiences designed to help people discover businesses more easily, while declining to provide more details about the experiment.

The company has emphasized that such tests are routine and not all will become widely available. Still, the Flipkart experiment fits a broader strategy Google has been outlining for months.

Why the Flipkart partnership matters

Google’s relationship with Flipkart gives the test added significance. The company is not only working with Flipkart as a technology partner, but also as an investor. Google put about $350 million into Flipkart in 2024 as part of a financing round led by the U.S. retailer, taking a minority stake in the e-commerce company.

That dual role makes Flipkart a natural candidate for early experiments in AI commerce. It also shows how large platforms are building intertwined commercial and technical relationships as they try to define the next generation of shopping tools.

India is a particularly strategic market for this kind of testing. The country is the world’s second-largest internet market, with more than a billion internet subscribers, and its e-commerce sector is highly competitive. Flipkart and Amazon dominate much of the conversation around online shopping, and that rivalry intensifies during the festive season when retailers launch some of their biggest promotional campaigns of the year.

Why festive season timing is crucial

The timing of Google’s test is unlikely to be accidental. Indian consumers typically increase spending around major holidays, and e-commerce companies prepare extensive sale events to capture that demand. If Google can make product discovery and checkout feel seamless in Gemini and AI Mode, it may increase the chances that users finish a purchase without abandoning the journey.

That matters for retailers as well. A direct-buy button inside an AI interface can reduce friction, shorten decision time, and potentially lift conversions. It also gives Google more influence over how users move through the shopping funnel.

How is Google’s commerce strategy changing?

Google is evolving from a search engine that helps people find products into a platform that may increasingly help them purchase those products. The company has been building the technical infrastructure to support that shift through its Universal Commerce Protocol, or UCP.

What is Universal Commerce Protocol?

Universal Commerce Protocol is Google’s open standard for letting AI agents interact with retailers across the shopping journey, including checkout. Google introduced it earlier this year as a way to enable purchases through Gemini and AI Mode, initially using a Google-hosted checkout experience.

Since then, Google has expanded UCP to support additional functions, including the ability for shoppers to transfer items to a retailer’s own site to complete the purchase. That flexibility suggests the company is trying to accommodate different retail setups instead of imposing a single transaction model.

The Flipkart test appears to sit alongside that work rather than replace it. The key distinction is that the checkout flow seen in India is branded by Flipkart, not Google, and the mechanism behind it has not been publicly detailed.

What makes this different from Google’s earlier demos?

This test is different because it looks less like a fully Google-controlled checkout and more like an embedded retailer checkout experience. In the earlier UCP demonstration, Google showed shoppers being able to buy eligible products through a Google-hosted flow. In the Flipkart test, the retailer’s own checkout branding appears once the Buy button is tapped.

That suggests Google may be experimenting with multiple transaction models. One could prioritize Google-managed commerce infrastructure, while another could give retailers more direct control over the final steps of the purchase.

For consumers, the main appeal is convenience. For Google and its partners, the bigger issue is who controls the transaction, the data, and the customer relationship.

Which products are included in the early test?

The test is limited to a small product group. According to people familiar with the matter, it currently covers select smartphones, electronics, and mobile accessories. Not every user can see the Buy button, and not every Flipkart listing is eligible.

Other users who open Gemini or AI Mode may still see ordinary Flipkart product listings without any direct-buy option. That uneven rollout is typical of early Google experiments, which often begin with a narrow audience before expanding if the company is satisfied with performance.

• Smartphones
• Electronics
• Mobile accessories

How does this affect competitors like Amazon?

It does not yet affect them directly in a functional sense, but it could reshape the competitive dynamics of AI-assisted shopping over time. Amazon listings were visible in the same Gemini and AI Mode experience, but the Buy button was not shown on those rival listings in the test seen by TechCrunch.

That matters because a direct-purchase option can steer consumer behavior. If one retailer is easier to buy from inside an AI interface than another, even small differences in friction can influence which merchant wins the sale.

In India, where Amazon and Flipkart are locked in a long-running battle for online shoppers, even modest interface advantages can have outsized importance during peak shopping periods.

Why is Google racing to add shopping features to AI?

Google is racing to add shopping features because AI products are becoming a new front door for online discovery, and the company does not want those experiences to stop at answering questions. If users increasingly begin shopping journeys inside chatbots or AI search tools, the platform that helps complete the transaction stands to gain influence and potential revenue.

The move also responds to pressure from competitors. OpenAI and others have been adding commerce-related capabilities to their own AI offerings, raising the possibility that shopping will become a standard feature of consumer AI products.

In that environment, a recommendation engine is no longer enough. The competitive prize is to become the place where users not only ask what to buy, but actually buy it.

Google’s public stance is that it is still experimenting, but the broader industry trend is clear: AI systems are shifting from advisory tools into transaction platforms.

What comes next for the India rollout?

Google is expected to broaden the test later in October, ahead of India’s festive shopping season, according to one of the people familiar with the plan. If the rollout expands as expected, more users may begin seeing the Buy button on additional Flipkart listings inside Gemini and AI Mode.

Even if the test remains limited, it will be watched closely. Retailers, rivals, and AI developers all have reasons to care about whether Google can translate AI product discovery into measurable checkout behavior.

For Google, the experiment is part of a larger question: can an AI assistant become a trustworthy shopping layer without losing the speed and simplicity that made it useful in the first place?

Timeline of Google’s AI commerce push

Google’s Flipkart test did not emerge in isolation. It is the latest step in a broader sequence of commercial AI efforts.

What this means for shoppers and retailers

For shoppers, the immediate benefit is convenience. A direct-buy button can reduce steps, especially for people already using AI to compare products or narrow choices. It may be particularly useful on mobile devices, where jumping between apps and browsers can be cumbersome.

For retailers, the implications are more complex. On one hand, AI-powered shopping surfaces can deliver more qualified traffic and higher purchase intent. On the other, retailers may have to accept a new layer of platform control over the discovery process.

There is also a strategic tradeoff. If AI assistants become the primary place where people begin shopping, retailers may have to compete not just on price and logistics but on how well their products are surfaced by the assistant itself.

What Google and Flipkart have not said

Several important details remain unclear. Google has not explained the technology behind the test, how checkout data will be handled, whether the flow uses UCP, or how broadly the feature will eventually roll out. Flipkart also did not immediately respond to requests for comment.

Those unanswered questions matter because they determine how much control Google retains over the shopping experience and how much freedom retailers have inside it. The answer will likely shape the long-term economics of AI commerce in one of the world’s most important digital markets.

For now, the test is small, experimental, and limited to a narrow set of users and products. But it is also a clear signal: Google wants its AI products to do more than talk. It wants them to sell.

Bottom line

Google’s India experiment with Flipkart shows how quickly AI interfaces are evolving from information tools into shopping environments. If the test expands as planned, it could become one of the clearest examples yet of AI-driven commerce entering mainstream consumer behavior in a major market.

That would mark a notable step for Google, a company whose core business has long depended on helping people find things online. Now it is trying to help them buy those things too.