Gemini has put a "Buy" button right inside the AI chat in India. On September 27, 2026, TechCrunch first reported that Google is testing a new feature on a small scale in India: select Flipkart product listings inside Gemini and Search's AI Mode now show a "Buy" button, which takes users straight into a Flipkart-branded checkout flow without ever leaving the AI interface. A Google spokesperson responded that the company is "always testing new features to help people discover businesses," but declined to share further details.
What the test looks like
Only some users in India can see the button for now, and product coverage is limited, concentrated on smartphones, electronics, and mobile accessories. An interesting detail: in the same interface seen by TechCrunch, Amazon listings also appeared — but without the same in-interface purchase treatment. People familiar with the matter say Google plans to roll the experience out more broadly in October, just ahead of India's Diwali festive shopping season.
Why India and Flipkart
India is the world's second-largest internet market, and the e-commerce sales around Diwali are the year's biggest consumer moment, so testing here is no surprise. Google's relationship with Flipkart was also laid down well in advance: in 2024, Google invested about $350 million in the Indian e-commerce giant as part of a Walmart-led funding round. Back in April, Gemini's shopping tools had already launched in India, letting users compare prices and browse product listings in chat — but purchases still required jumping to an external site. This test closes that "last mile" inside the AI interface.
AI moves from "finding it for you" to "buying it for you"
The backdrop is the open standard Google introduced earlier this year, the Universal Commerce Protocol (UCP), which lets AI assistants speak a common language with retailers' inventory and checkout systems. From product discovery to closed-loop transactions, Google is rewriting the monetization logic of AI search: the entry point no longer just routes traffic, it directly participates in the sale. The Gemini product line has been moving aggressively in this direction — from agentic video understanding (Gemini launches agentic video understanding: long videos no longer watched frame by frame) to in-chat purchasing, the throughline is turning AI from "answering questions" into "getting things done." For e-commerce platforms, the question becomes: when users stop opening apps and comparison and checkout both happen in a chat window, who sets the rules of traffic?
