Google Testing Agentic Shopping and Native Checkout Features in Gemini

Google appears to be recently working on adding a new Shopping Cart item inside the settings menu of the Gemini app. The same build also shows a redesign that moves existing items like Gems and Scheduled Actions from the sidebar into a new settings page.

This Shopping Cart signals an intent to let users browse, select, and purchase products directly within Gemini. In other words, rather than acting as a helper that sends users off to external shopping sites, the direction is to grow Gemini into a self-contained commerce surface.

This move ties into Google's already ongoing agentic shopping strategy. Through the Universal Commerce Protocol announced at NRF in January 2026, Google began supporting checkout at retailers such as Target, Gap, Etsy, and Wayfair, and a dedicated cart inside Gemini could provide a persistent home for these checkout flows.

In particular, Skills for Gemini in Chrome, which surfaced just days ago, points in the same direction. This feature offers reusable, one-click prompt workflows that work across tabs and pages, and when viewed together with the previously spotted Agent tab with Skills, it suggests that Google is trying to unify the Chrome-based Gemini experience with the main Gemini UI.

If Chrome Skills, automated browsing, and agentic shopping converge into a single interface, Gemini in Chrome could function like a desktop-class AI app that handles browsing, automation, and commerce all in one place. That would put it in direct competition with ChatGPT and Copilot in shopping, and with AI browsers like OpenAI Atlas and Perplexity Comet in agentic browsing.

However, all of these features are still unreleased and under development. The settings overhaul itself has not been officially announced either, but with Google I/O scheduled for May 19-20, plans related to shopping, Skills, and Chrome integration are likely to be revealed there.