Walmart is preparing for a future in which AI agents will do the shopping.

Big picture: AI shopping agents are poised to reshape retail in ways that have only just begun to be seen. Walmart’s actions in this new frontier highlight a crucial truth: The race to win AI shoppers is only just beginning and rules are being written as they go.

Walmart prepares for a major shift in retail as artificial intelligence agents change not only how people shop but also who or what is doing the shopping. The retail giant is looking at ways to make their products appealing to not only human customers but also to AI bots who may soon be making purchases on behalf of consumers. Hari Vasudev is Walmart’s US Chief Technology Officer. “Advertising will have to evolve.” AI’s influence on how consumers research products is already visible, with search engines favoring AI-generated summaries rather than traditional links. The next phase, where AI agents make purchases independently, promises to be a much greater disruption. Imagine a future in which a user tells OpenAI Operator to “Restock my pantry” “Find a 65-inch TV.” and the agent scans the internet, evaluates options according to the user’s preferences and history, and completes a purchase, including payment. Robert Hetu is a retail analyst with Gartner. He said that retailers will have to rethink everything, from product descriptions and pricing strategies. If third-party agents are handling checkout, retailers run the risk of losing direct relationships with their customers, reducing brand loyalty, and losing access to valuable information.

Walmart doesn’t wait passively. It is developing its own AI-based shopping tools that are accessible through its app and website. These tools can handle tasks as mundane as weekly grocery reorders, or as whimsical as “unicorn-themed party for my daughter,” Vasudev pointed out. The retailer’s online sales grew by 22 percent in the last quarter. However, it is also preparing to a scenario where customers prefer third-party agents such as Operator.

Vasudev believes that an industry-wide protocol, allowing third party agents to communicate with retailer’s systems, will emerge. These standards, which are still being developed, will allow retailers to share tailored product recommendations. Vasudev explained that agents could scan sites independently in a similar way to a shopper browsing an entire store without assistance.

But retailers face challenges. Agents such as Operator give priority to factors like search rankings, paid ads, and sponsored posts when surfacing products. This algorithmic approach is in stark contrast to human behavior, even though user prompts and preferences are still central. Hetu explained that bots will not linger on emotionally charged images, and retailers need to make lightning-fast decisions about pricing by offering discounts, or risk instant price comparisons from agents. Hetu tempers the hype despite the buzz. Over 80 percent still shop in physical stores and AI-driven sales remain a tiny fraction of sales. “This is going to take time to transform,” He said.