🔍 Read the full analysis: What Amazon’s AI Agent Block And Walmart’s Welcome Say About Retail on ThorstenMeyerAI.com
Get everyday essentials 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
Walmart announced an Agent Gateway and customer-controlled Agent Profiles on Oct. 21, while Amazon has blocked Meta’s Muse shopping agent from accessing its site. The moves show retailers taking different approaches to AI shopping, as control over customer data, checkout and sales attribution remains unsettled.
Walmart has opened a structured route for approved AI shopping agents to search and buy products, while Amazon has blocked Meta’s Muse from accessing its website, setting up a dispute over how automated systems interact with retailers. Walmart announced its Agent Gateway and customer-controlled Agent Profiles on Oct. 21; Amazon says outside agents must identify themselves and comply with its site terms.
Walmart’s Agent Gateway uses the Model Context Protocol (MCP) to let approved agents access retail functions without navigating the store’s ordinary consumer pages. Its Agent Profiles are designed to give shoppers controls such as spending limits, according to the source report. The announcement does not mean every agent can automatically shop at Walmart; access depends on the retailer’s system and the agent’s authorization.
Amazon’s approach is different. When Meta’s Muse attempted to shop on Amazon, users saw a notice saying continued access by an unauthorized AI agent violated Amazon’s Conditions of Use. Amazon said Meta had not told it Muse would shop on the site, argued that the agent did not identify itself and said it appeared to capture and store customer credentials. Meta’s launch post said Muse had no visibility into users’ passwords or payment methods; the competing descriptions have not been independently resolved in the source material.
The moves follow Meta’s Sept. 8 launch of Muse, an agent that can use a browser to research products and fill out forms. Meta said on its Oct. 21 earnings call that more than 10 million people in the United States and Canada had set up Muse in roughly six weeks, shopping was its most common use and the average purchase exceeded $200. Those are Meta’s figures, not independently audited measures of completed agent purchases.
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 to Control Agent Checkout
The dispute concerns more than whether an automated shopper can load a product page. Retailers want control over identity, access and checkout; agent companies want to make it easier for customers to compare offers and complete purchases across stores. If agents become a regular shopping interface, the company that controls the agent may shape what consumers see, where they buy and which seller receives the order.
There is also a commercial question about who gets credit for a sale. The source report says Amazon earned more than $68 billion in advertising revenue last year, citing GeekWire, and quotes Meta chief executive Mark Zuckerberg saying the company expects to earn “a small fee” from transactions. Those figures and statements point to competing business interests, but they do not establish how much revenue agent shopping will generate or how it will be divided.
For shoppers, an agent can reduce the work of finding and comparing products, but convenience does not settle practical questions. A buyer still needs to know which retailer is responsible, what return policy applies and whether a purchase requires approval. Spending caps and payment controls matter because an agent may be able to act on the user’s behalf, while the exact permissions and safeguards differ by service.
Products Worth Considering
Sonos Era 100 Wireless Smart Speaker, Black
Sonos Era 100 SL Compact Microphone-Free Speaker, Black
Amazon Kindle Scribe Colorsoft 64GB (1st Generation)
As an affiliate, we earn on qualifying purchases.
From Browser Bots to Retail Gateways
There are three main approaches described in the source report. A browser-based agent, such as Muse, visits websites much as a person would, potentially using a cloud computer and user-authorized access. This can work across many sites, but retailers may have difficulty distinguishing automated activity from a shopper and may object to how credentials or site rules are handled.
A second approach is a retailer’s own approved connection. Walmart’s gateway and Shopify’s tools are examples of structured access, where a service exposes specified shopping functions rather than asking an agent to imitate human clicks. Shopify introduced tools for agents to get, update and complete a checkout in September, according to the report.
Companies are also working on shared protocols. Meta, Sierra, Walmart, Shopify and Stripe announced the Personal Agent Protocol, also called Poppy, in October; its draft describes how agents can identify themselves and carry customer permissions across interactions. Other efforts include the Universal Commerce Protocol, backed by Google and Shopify, and the Agentic Commerce Protocol from OpenAI and Stripe. The protocols are at different stages and are not proof that agents can shop seamlessly across retailers.
Earlier product developments show the gap between building a checkout connection and establishing a shopping habit. OpenAI introduced Instant Checkout in ChatGPT using its protocol in September 2025, then said in March 2026 it was letting merchants use their own checkout experiences while focusing on product discovery. The source report cites HUMAN Security data for September showing 81% of agent requests went to product and search pages and 2.4% to checkout and payment. That is a measure of requests in that dataset and period, not a general estimate of all agent shopping.
“Continued access by an unauthorized AI agent violates Amazon’s Conditions of Use, to which our customers have agreed.”
— Amazon, in the notice shown to Muse users, as reported by ThorstenMeyerAI.com
Access Rules and Sales Attribution Remain Open
It is not yet clear how many purchases Muse or other agents complete, how often retailers approve agent access, or whether the announced tools will be adopted widely. Meta’s setup and purchase figures are company-reported, and the source does not provide independent verification or a comparison baseline for them.
The dispute over Muse’s handling of credentials also remains unresolved in the material provided: Amazon’s concern and Meta’s denial are both attributable statements, not a settled technical finding. The report does not specify how Walmart will review agents for gateway access, what protections apply if an agent makes an unwanted purchase, or how responsibility will be divided among a shopper, agent provider and retailer.
Sales attribution is another open issue. Traditional affiliate links can record a click that leads to a purchase; an agent that reads product information and buys through another route may leave no equivalent referral record. The source describes that as a risk for publishers, but supplies no measured estimate of lost commissions or a settled industry-wide payment model.
Watch Agent Access and Checkout Adoption
The next test is whether retailers and agent providers move from announcements and draft protocols to working, widely available checkout. Walmart’s gateway and profiles will need to show which agents are eligible, what shoppers can authorize and how purchasing controls operate in practice. Amazon’s restrictions may also prompt further negotiations or technical changes, though the source does not report a resolution.
Protocol development and merchant adoption are worth watching alongside user growth. New integrations could make product discovery easier without changing who processes payment or handles returns. For consumers, the practical questions remain whether purchases require confirmation, whether spending limits can be set, and which company is responsible when an order, payment or return goes wrong.
Key Questions
What did Walmart announce?
Walmart announced an Agent Gateway using MCP and Agent Profiles intended to give shoppers controls over an AI agent’s activity, including spending limits, according to the source report. The announcement does not establish that all shopping agents already have access.
Why did Amazon block Meta’s Muse?
Amazon said Muse was an unauthorized agent that had not identified itself and appeared to capture and store customer credentials. Meta said Muse could not see users’ passwords or payment methods. The source does not independently settle the disagreement.
Can AI agents already complete purchases?
Some systems and merchant tools are designed to support checkout, but availability varies. The source cites September data in which 2.4% of agent requests in HUMAN Security’s dataset went to checkout and payment; that figure does not describe all shopping agents or all transactions.
What should shoppers check before letting an agent buy?
Check whether purchases require approval, set a spending cap if available, and review the payment method, retailer and applicable return policy. The controls and responsibility for a purchase can differ across agent and retailer services.
Source: ThorstenMeyerAI.com
