Retail is entering a new phase of automation in which chatbots are shifting from answering questions to actively steering shopping decisions—and in some cases completing transactions.

Walmart is leaning into that shift by plugging its product catalogue and customer accounts into conversational tools designed to shorten the path from “I need this” to “it’s on the way,” while Amazon is moving more cautiously, keeping more of the shopping journey inside its own ecosystem.

At the center of the change is a simple idea: a chatbot that can interpret intent, search across large catalogues, summarize reviews, compare options, and build a cart can become the front door to shopping—potentially replacing the traditional sequence of browsing, searching, filtering, and clicking through product pages.

That re-routes not only how consumers discover products, but also who controls the data, the customer relationship, and the advertising value that typically sits on a retailer’s website or app.

Walmart has been expanding its in-house shopping assistant, including its “Sparky” chatbot, and has also moved to integrate with Google’s Gemini so shoppers can browse and buy without leaving the chat interface.

That approach treats conversational AI as a new storefront: the assistant becomes the shopping surface, while Walmart provides assortment, fulfillment, pricing, and account-linked personalization when a customer connects their Walmart or Sam’s Club profile.

The technical direction matters because it changes bargaining power.

If a chatbot can route demand across multiple retailers, the assistant becomes an intermediary that can influence what customers see first—much like search engines did for web traffic—while retailers risk becoming interchangeable fulfillment backends.

Walmart’s strategy suggests it would rather be early and shape the rules of that interface than wait for a third party to capture the customer relationship.

Amazon, by contrast, has built its own AI shopping assistant, Rufus, embedded in its app and website, trained on its product catalogue and other shopping information.

The advantage is obvious: the guidance, the ad placement, and the purchase all happen on Amazon-controlled real estate, keeping the feedback loops—clicks, conversions, and customer signals—inside the platform.

But the same forces that make chat shopping powerful also create threats for retailers and platforms alike.

If an AI agent can shop “on behalf” of the user, it can bypass the familiar advertising and ranking systems that have long shaped e-commerce economics, and it can compress product differentiation into a short set of chatbot answers.

That puts new pressure on how recommendations are generated and whether they are influenced by advertising incentives or commercial partnerships.

There are also practical constraints that will determine how quickly conversational commerce scales.

Payments and identity verification have to be smooth enough for a one-chat checkout experience, fraud controls have to adapt to automated purchasing, and consumers have to trust that the assistant is not “hallucinating” product claims or misreading inventory and delivery options.

Retailers are also wary of exposing too much catalogue and pricing intelligence to systems they do not fully control.

The contrast between Walmart and Amazon is therefore less about who has better AI and more about who owns the shopping layer.

Walmart is betting that partnering into a conversational interface can grow reach and keep pace with changing consumer behavior, even if it means sharing the front end with a tech platform.

Amazon is betting that keeping the assistant native to Amazon protects its core economics, while still capturing the efficiency gains of AI-led discovery and guidance.

What to watch next:
- Whether “agent-led” checkout inside chat apps expands beyond the United States and to more payment methods
- How retailers and platforms disclose whether chatbot recommendations are influenced by advertising or commercial placement
- Whether consumers adopt automated cart-building and auto-buy features at scale, or prefer “assistive” chat that stops short of purchase
- How fraud prevention and returns policies evolve when purchases are initiated by AI assistants
- Whether brands shift budgets from traditional search optimization toward formats tailored for AI shopping assistants