Walmart is rolling out controls for AI agents that shop on a customer's behalf across Walmart and Sam's Club. Agent Profiles let shoppers authorize an agent, set a spending limit such as $50, $100, or $200, and see which orders an agent placed. A new Agent Gateway built on the Model Context Protocol (MCP) gives agents a direct connection for discovery, add to cart, purchasing, and order tracking, and new rules keep agents away from payment details and unauthorized data. For eCommerce brands, this means a growing share of orders will be placed by software that reads your catalog, compares your price, and checks your delivery promise without ever seeing your marketing. The hires that matter most are an agentic commerce owner, a marketplace or digital shelf manager who keeps catalog data agent-ready, an eCommerce technology lead fluent in protocols like MCP, a pricing and fulfillment owner, and an analyst who can measure agent-driven sales.
Digital Commerce 360 reported on October 23 that Walmart is introducing tools for shoppers to supervise AI agents that buy on their behalf. Walmart announced the changes on October 21, and they cover three pieces: Agent Profiles, an Agent Gateway, and a new set of purchasing rules for agents.
The timing matters. Walmart already sells through its own assistant, Sparky, and through integrations with ChatGPT, Gemini, and Meta's Muse. Until now, though, a purchase placed by an agent looked exactly like a purchase placed by the customer. Walmart is now building the plumbing to tell them apart, set limits on them, and let agents connect directly instead of clicking through pages built for people. Daniel Danker, Walmart's executive vice president of AI acceleration, product and design, framed it as Walmart "setting expectations for how AI agents work with customers and our company."
For the brands and sellers whose products show up in those agent conversations, that is a hiring signal.
What Walmart Announced
| Update | What It Does | Status |
|---|---|---|
| Agent Profiles | Shoppers authorize agents, set spending limits ($50, $100, $200, or approve every order), track agent purchases, and delete profiles anytime | Launching within the next month |
| Agent Gateway | A direct MCP connection for discovery, add to cart, purchasing, and order tracking | Limited beta for companies and developers |
| Purchasing rules | Agents need approval by default, cannot see card details, cannot change payment methods, and are blocked if they break the terms | Part of the rollout |
A few details are worth noting. Agents on Walmart's site see the same prices a human would, with offers and fees shown before purchase. Unintended agent purchases go through Walmart's normal returns process. And the gateway can help an agent find a shopper's usual items, like a preferred brand of milk, without handing over the customer's full shopping history. Walmart has not said which agents will support profiles at launch.
Why Spending Limits Matter for Brands
That comes from an August NMI survey of 1,000 U.S. adults cited in the article, which also found 11% had already used AI to complete a transaction. Spending limits and approval steps are Walmart's answer to that trust gap. They make shoppers more willing to delegate, which means more orders where an agent, not a person, is doing the browsing.
When an agent shops, the things that usually win a sale on a product page, such as lifestyle imagery, brand storytelling, and on-site promotions, carry far less weight. The agent works from structured data: what the product is, what it costs, whether it is in stock, and when it will arrive. If that data is incomplete or wrong, the agent recommends something else, and the shopper may never know your product was considered.
Walmart's own marketplace leadership pointed in the same direction. Manish Joneja, senior vice president of Walmart Global Marketplace and Walmart Fulfillment Services, said AI-driven discovery could help brands with less name recognition reach customers, but that accurate product information, competitive pricing, availability, and dependable delivery promises will matter more. Danker noted earlier this year that vitamins and protein supplements were the top two product types bought from Walmart through ChatGPT, often starting from a broad conversation rather than a product search.
Agentic commerce does not reward the loudest brand. It rewards the brand whose data an agent can trust. That makes catalog quality, pricing, and delivery accuracy hiring problems, not just tooling problems.
Mapping Walmart's Agent Controls to the Hires
| What Changes | Function It Requires | Who Typically Owns It |
|---|---|---|
| Agents shop across Walmart, ChatGPT, Gemini, and Muse | Cross-platform agentic commerce strategy | Director or VP of eCommerce |
| Agents choose products from catalog data | Listing quality, attributes, content accuracy | Marketplace Manager or Digital Shelf Manager |
| Agents connect through MCP gateways | Protocol fluency and integrations | eCommerce Technology Lead |
| Agents compare price and delivery promises | Price consistency, inventory, fulfillment accuracy | Pricing Lead and Fulfillment Lead |
| Agent orders are now tracked separately | Attribution and performance reporting | eCommerce Analyst |
The Roles to Prioritize
An Agentic Commerce Owner
Owns: strategy across Walmart, ChatGPT, Gemini, and other agent platformsWalmart's move follows Google's Universal Cart and TikTok's AI shopping assistant by only a few weeks. Each platform has its own rules, data requirements, and checkout model. Someone at the Director of eCommerce level or above needs to decide which agent channels matter, set catalog and pricing standards across them, and define how success is measured. Without that owner, agent readiness becomes a side project split between marketplace, IT, and an agency. Our explainer on what agentic commerce is covers the broader shift.
Marketplace and Digital Shelf Manager
Owns: Walmart listings, attributes, content accuracy, catalog healthWalmart is actively working with sellers to improve listings so AI tools can understand them. That is the core of the Digital Shelf Manager role, now judged by a reader that never skims. In interviews, ask candidates how they would audit a Walmart Marketplace catalog for an AI agent rather than a human shopper. Strong candidates talk about attribute completeness, consistent variant data, and how a product would be described in a conversational query, not just keyword placement. Our guide to the biggest global eCommerce marketplaces covers how this role varies by platform.
eCommerce Technology Lead
Owns: protocol integrations such as MCP and Google's UCP, data flow between systemsWalmart built its Agent Gateway on the Model Context Protocol, the same open standard many AI applications use to connect to outside tools. Google's checkout runs on its Universal Commerce Protocol. Brands do not need to build gateways of their own tomorrow, but they do need someone who understands how agents connect, what data flows through those connections, and when an integration is worth the effort. That is an eCommerce technology leader, or at a Shopify brand, a senior Shopify developer who has worked on integrations, not just storefronts.
Pricing and Fulfillment Owners
Owns: price consistency, inventory accuracy, delivery promisesAgents see the same prices people see, and they compare them instantly. Joneja specifically named competitive pricing, availability, and reliable delivery promises as factors that will matter more. A delivery promise that slips, or a price that is higher on Walmart than on your own site, is the kind of signal an agent is built to catch. Brands need a clear owner for pricing consistency and an eCommerce fulfillment and logistics lead who treats delivery estimates as customer-facing data, especially heading into the holidays.
eCommerce Analyst
Owns: measuring agent-driven sales and returnsOne of the quiet changes in Walmart's announcement is that agent purchases will no longer blend in with customer orders. That opens the door to real measurement: which products agents choose, how often agent orders get returned, and whether agent sales grow or cannibalize other channels. An eCommerce analyst who can build that reporting will give leadership something most teams do not have yet. We covered the attribution side in how AI recommendations are reshaping sales and attribution.
How This Fits the Bigger Agentic Commerce Picture
In the last month alone, Google rolled out Universal Cart, TikTok added an AI shopping assistant with one-click checkout, and now Walmart has set the ground rules for outside agents shopping its site. The pattern is consistent: platforms are making it easier and safer for shoppers to let AI buy for them, and the work of winning those orders is moving into data, integrations, pricing, and measurement.
Digital Commerce 360 noted that Walmart ranks No. 2 in its Top 1000 database of North America's largest online retailers, but No. 37 in its AI Commerce Rankings with ReFiBuy. Even the biggest players are still building readiness. For most brands, the gap is not a lack of interest in AI. It is a lack of people whose job is to make the brand legible to it.
What to Do Before Black Friday
- Audit your Walmart catalog for an agent, not a person. Check attributes, variants, and descriptions for completeness and consistency.
- Check price parity. Agents see the same prices shoppers do. Any gap between Walmart, Amazon, and your own site is now easier to spot.
- Tighten delivery promises. Make sure the delivery estimates you show are ones you can keep through peak season.
- Plan for agent reporting. Once agent orders are labeled, decide who will track them and what success looks like.
- Name an owner. If no single person is accountable for how your brand performs with AI shopping agents, start that search now. Our eCommerce hiring timeline shows how long these searches typically take.
Frequently Asked Questions
What are Walmart Agent Profiles?
Agent Profiles are a Walmart feature, planned to launch within a month of its October 2026 announcement, that lets shoppers authorize third-party AI agents, set spending limits, and see which purchases an agent made. Shoppers can approve every agent purchase individually or let an agent spend up to a preset limit such as $50, $100, or $200, and they can delete a profile at any time. The feature applies across Walmart and Sam's Club.
What is the Walmart Agent Gateway?
The Agent Gateway gives AI agents a direct connection to Walmart's shopping services instead of making them click through a website built for people. It uses the Model Context Protocol (MCP), an open standard developed by Anthropic, and supports product discovery, add to cart, purchasing, and order tracking. Walmart said it will launch in a limited beta for companies and developers.
What rules does Walmart place on AI shopping agents?
By default, shoppers must approve agent purchases. Agents cannot act beyond the shopper's instructions, access information they are not authorized to see, receive payment card details, or add or change payment methods. Agents see the same prices a human would, unintended purchases go through Walmart's normal returns process, and agents that break the rules are blocked.
How does AI agent shopping affect Walmart Marketplace sellers?
Walmart's marketplace leadership has said AI-driven discovery could help lesser-known brands reach customers, but that accurate product information, competitive prices, availability, and reliable delivery promises will matter more. Walmart is working with sellers to improve listings and catalog quality so AI tools can understand and recommend their products.
What roles should eCommerce brands hire for AI agent shopping?
The most important roles are an agentic commerce owner at the Director of eCommerce level or above, a Marketplace or Digital Shelf Manager who keeps catalog data agent-ready, an eCommerce Technology Lead who understands protocols like MCP, a pricing and fulfillment owner who keeps prices and delivery promises accurate, and an eCommerce analyst who can separate and measure agent-driven orders.
Source: Beth Duckett, "Walmart unveils AI agent shopping controls, spending limits," Digital Commerce 360, October 23, 2026.