Walmart has announced plans to permit AI agents to conduct purchases on its website, positioning itself differently from competitors in the emerging AI shopping space. The initiative will begin the following month and will include multiple safeguards designed to keep customers in control of their spending and purchasing decisions.
The retailer has established specific operational guidelines that agents must follow. These requirements include restrictions on taking actions outside the customer’s stated intentions and prohibitions on modifying payment methods. Agents will also be limited to accessing only the information necessary to complete requested tasks. Walmart indicated it will enforce compliance with these terms by blocking agents that violate them.
The shopping feature will operate through Agent Profiles, a new section within Walmart’s platform. The system defaults to requiring customer approval for each purchase, though users can alternatively set spending limits, such as capping AI purchases at $100. Customers will be able to review agent purchases in a separate section from their own orders and can disable the agent feature at any time.
Walmart has developed an Agent Gateway that integrates AI agents using the Model Context Protocol, an open technical standard enabling AI systems to connect with various online services. This technical approach reportedly prevents agents from encountering navigation difficulties and restricts access to broader customer data. The company stated that while agents can locate frequently purchased items like specific bread brands or milk types, the customer’s complete shopping history remains protected from the agents. The retailer also plans to extend this capability to Sam’s Club, its warehouse club subsidiary.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI