🔍 Read the full analysis: AI Shopping Agents: Why Amazon Blocks Them And Walmart Allows Them on ThorstenMeyerAI.com
Get the little things that make your day delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Meta’s Muse shopping agent cannot access Amazon, which says outside agents must identify themselves and comply with its rules. Walmart is taking a different approach, announcing an Agent Gateway and profiles that let approved agents search and purchase under defined permissions. The disagreement reflects competing interests in security, customer relationships and the economics of online shopping.
Amazon has blocked Meta’s Muse shopping agent from accessing its site, while Walmart is building an Agent Gateway that lets approved AI agents search products and make purchases. The contrasting moves put retailers on different sides of a fast-growing dispute over who may shop on their platforms, how purchases are authorized and which company controls the customer relationship.
The immediate conflict surfaced after Muse tried to compare prices. In an account published by The Verge, Meta’s agent listed Sam’s Club, Walmart and Costco as sources for granola bars, then said it could not check Amazon because the retailer had blocked it. Amazon’s restriction means an agent using the browser like a person cannot reliably compare prices across all major retailers.
Amazon said Muse had not identified itself as an agent and that Meta had not told the company it would shop on Amazon. Amazon also said Muse appeared to capture and store customer credentials. Meta’s launch post, in contrast, said Muse has no visibility into users’ passwords or payment methods. Those statements describe competing accounts of the agent’s operation; the available source material does not independently resolve the disagreement.
Walmart’s October 21 announcement took a more structured approach. Its Agent Gateway is designed to let approved agents access shopping functions, while Agent Profiles give customers controls such as spending limits. The announcement followed Meta’s September disclosure of retail connectors that included Walmart, Best Buy, Gap, Sephora and Wayfair. Meta said it expected to earn a small fee from transactions made through its agent.
Agentic shopping
Three ways an AI agent buys for you
Every fight between shops and AI agents comes down to which of these three routes the agent uses, and whether the shop agreed to it.
The browser, like a human
The agent opens pages built for people, signs in with your login and clicks through checkout. Works almost anywhere, which is why shops object: they can't tell the agent from you.
Examples: Meta's Muse (from a virtual machine in Meta's cloud), Perplexity's Comet (from your own computer)
The shop's own agent door
Approved agents search and buy through a structured interface instead of clicking through pages. The shop sets the rules.
Examples: Walmart Agent Gateway (MCP, beta), Shopify MCP server and WebMCP checkout tools
Shared protocols
Common rules for who the agent is, what you allowed, and how the checkout and payment are confirmed.
Examples: Personal Agent Protocol (draft), Universal Commerce Protocol, Agentic Commerce Protocol
The protocols competing to be the standard
| Protocol | Backers | Covers | Status |
|---|---|---|---|
| Personal Agent Protocol (Poppy) | Sierra, Meta, Walmart, Shopify, Stripe + 35 design partners | Agent identity, customer permission, one session across site, API and company agent | Draft 0.1 (9 Oct 2026) |
| Universal Commerce Protocol (UCP) | Google, Shopify; site also lists Etsy, Wayfair, Target, Walmart, Amazon, Microsoft | Discovery, cart, checkout, orders; payment mandates via AP2 | Published |
| Agentic Commerce Protocol (ACP) | OpenAI, Stripe | Checkout between agent and merchant | Open source; ChatGPT's own checkout scaled back in March 2026 |
What a careful agent purchase looks like
- You give the task and a limit, for example a $50 cap.
- The agent identifies itself to the shop and gets only the access you approved.
- It prepares the cart and shows you an approval card.
- Payment goes through a token or one-time card, so the agent never sees your card.
- The shop’s normal returns apply if the agent got it wrong.
Sources: Meta (8 Sep 2026), Walmart (21 Oct 2026), TechCrunch on Shopify (28 Sep 2026), Sierra (9 Oct 2026), ucp.dev, agenticcommerce.dev, OpenAI (24 Mar 2026). Purchase steps: Walmart’s agent terms and Meta’s Muse description.
Powered by Thorsten Meyer AI
Retailers Compete for Agent Access
Shopping agents can change how customers discover products and where they complete purchases. If an agent selects a retailer, the store may gain a sale without the customer visiting its website in the usual way. If the agent is blocked, a retailer may be absent from its comparisons. The dispute is not only about whether software can browse a site; it concerns control of product discovery, checkout and customer data.
There is also a financial question for retailers and publishers. Amazon reported more than $68 billion in advertising revenue last year, according to GeekWire, giving the company a significant stake in how shoppers find products on its platform. Meta CEO Mark Zuckerberg has said Meta expects to take a small fee from transactions. Meanwhile, affiliate publishers may miss commissions when an agent reads a review but does not follow a tracked link before buying. The source material describes that risk but does not quantify its effect.
For shoppers, convenience comes with practical questions about authorization and responsibility. A customer using an agent needs to know whether a purchase requires approval, how much the agent can spend and which retailer’s return rules apply. Walmart’s proposed profiles point toward permission controls, but the details of how those controls will work for customers and partner agents remain important.
Products Worth Considering
Withings Body Smart Wi-Fi and Bluetooth Body Composition Scale
Levoit Vital Pet Pro Smart Air Purifier for Pets, White
Blueair Blue Pure 511i Max Air Purifier for Home and Bedroom
As an affiliate, we earn on qualifying purchases.
Three Routes to an AI Purchase
Shopping agents can reach stores in different ways. One method is to use a browser as a person would: the agent opens pages, clicks links and may sign in with a customer’s credentials. Meta’s Muse runs in a virtual machine in Meta’s cloud. Retailers can object to this approach because browser activity may be difficult to distinguish from a human shopper, and it can raise questions about credentials, automation and site rules.
A second method is a retailer-approved connection. Walmart’s gateway uses the Model Context Protocol (MCP) to give approved agents access to shopping functions without having them click through ordinary web pages. Shopify has also developed structured tools for browser agents to manage checkout. A third approach is shared protocols intended to define how agents identify themselves, receive permission and interact with retailers. The source lists several efforts, including the Personal Agent Protocol and Google’s Universal Commerce Protocol, but their adoption and practical effects are still developing.
The timeline has moved quickly. Meta launched Muse on September 8, and Amazon restricted its access on September 20. Meta announced retail connectors on September 23. Walmart announced its Agent Gateway and Agent Profiles on October 21. Meta said on its October 21 earnings call that more than 10 million people in the United States and Canada had set up Muse in roughly six weeks, that shopping was its most common use and that the average purchase exceeded $200. These are Meta’s own figures, not independently verified usage data.
“Continued access by an unauthorized AI agent violates Amazon’s Conditions of Use, to which our customers have agreed.”
— Amazon, in a message shown to Muse users
Access Rules Still in Dispute
Several practical details remain unsettled. Amazon and Meta have offered different descriptions of Muse’s access and handling of credentials, and the source material does not include an independent technical examination that resolves those claims. It is also unclear how broadly Walmart’s gateway will be available, which agents will qualify for access, or what customer protections and dispute processes will accompany purchases made through it.
The protocols described in the source are at different stages, from published specifications to early drafts. Their presence does not establish that retailers or agents have adopted them widely. Nor do Meta’s reported setup and purchase figures show how often users complete transactions, how representative those users are or how much shopping shifts away from retailers’ own interfaces. A cited HUMAN Security analysis found that 81% of agent requests in its September data went to product and search pages, while 2.4% went to checkout and payment; those figures describe requests in that dataset, not all agent shopping.
Adoption Will Test the Models
The next test is whether retailers and agent providers can agree on access, identity and purchase permissions in a way that works across stores. Walmart’s gateway and spending profiles will need to move from an announcement to a usable service, while Amazon’s rules will shape whether outside agents can interact with its marketplace at all. Protocol efforts may provide shared tools, but their value will depend on merchants and agents putting them into operation.
Customers should check what an agent can do before connecting it to an account: whether each purchase needs approval, whether a spending cap is available, what payment method is used and which retailer handles returns. The competing company claims, early product announcements and limited activity data leave the broader outcome open. For now, the clearest difference is strategic: Amazon is restricting an outside agent, while Walmart is developing a controlled access route.
Key Questions
Why did Amazon block Meta’s Muse?
Amazon said Muse had not identified itself as an AI agent, that Meta had not told Amazon it would shop there, and that Muse appeared to capture and store customer credentials. Meta said Muse cannot see users’ passwords or payment methods. The available information does not independently settle the disagreement.
How is Walmart approaching AI shopping agents?
Walmart announced an Agent Gateway for approved agents and Agent Profiles that can include customer controls such as spending limits. The announcement describes its planned approach; the source does not establish how widely the service is available or which agents can use it.
Can AI agents currently complete purchases?
Some agent tools are designed to handle checkout, but the source reports that most agent requests in HUMAN Security’s September dataset went to product and search pages: 81%, compared with 2.4% to checkout and payment. Those figures apply to that dataset and do not establish the rate for all shopping agents.
What should shoppers check before using an agent?
Check whether the agent asks for approval before each purchase, offers a spending cap, uses a payment method you recognize and clearly identifies the retailer responsible for returns. Access and purchase terms may differ between retailers and agents.
Source: ThorstenMeyerAI.com
