🔍 Read the full analysis: Amazon Blocks AI Shopping Agents While Walmart Makes Room For Them on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
Amazon has blocked Meta’s Muse from shopping on its site, while Walmart is developing an Agent Gateway and profiles that let approved agents search and buy under customer-set limits. The competing approaches put control of shopping data, purchases and commissions at stake, though most agent activity still focuses on research rather than checkout.
Amazon has blocked Meta’s Muse from accessing its shopping site, while Walmart is opening a controlled route for AI agents to search its inventory and make purchases. The split shows retailers taking different approaches to software that shops on a customer’s behalf, with consequences for who controls the transaction and how shoppers, merchants and publishers are paid.
Meta launched Muse on September 8, describing it as a personal agent that can use a browser, fill in forms and act for a user. On September 20, users trying to reach Amazon through Muse saw a notice saying access by an unauthorized AI agent violated Amazon’s Conditions of Use. Amazon said Meta had not told it Muse would shop on its site, that the agent did not identify itself and that it appeared to capture and store customer credentials. Meta’s launch post said Muse could not see users’ passwords or payment methods.
Walmart announced its Agent Gateway and Agent Profiles on October 21. The gateway is designed to let approved agents search Walmart’s inventory and buy through an integration rather than by navigating ordinary web pages. Profiles can set spending limits for an agent. The announcement does not, by itself, establish how broadly the tools are available or how many customers are using them.
Meta reported 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. Meta said shopping was its most common use and the average purchase was over $200. Those numbers are company-reported figures; they do not show how often users completed purchases or how the average was calculated. A separate September HUMAN Security analysis cited in the source material found that 81% of agent requests went to product and search pages, while 2.4% went to checkout and payment pages.
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
Who Controls an Agent-Led Purchase
The disagreement is about more than whether an automated browser can open a product page. A shopping agent can influence what a customer sees, where they buy and which checkout they use. A retailer that supplies the agent’s access route can set conditions for product data, purchasing and customer permissions. A platform that owns the agent may instead become the shopper’s main point of contact.
That shift could also affect businesses that rely on referrals. The source material says Amazon Associates commissions are tied to customers using tracked links. If an agent reads a review but completes a purchase without using the link, a publisher may not receive credit. The scale of that potential effect is not established here. Meta chief executive Mark Zuckerberg said the company expects to earn “a small fee from transactions”, while the source material reports Amazon earned more than $68 billion from advertising last year, citing GeekWire. Those figures help explain why retailers and platforms have a financial interest in the route a purchase takes.
For shoppers, agent access can make product research and checkout easier, but it also raises questions about spending authorization, payment security and returns. A purchase made through an agent may involve the retailer’s policies, the agent provider’s terms and a payment service. Customers need to know which party is responsible when something goes wrong.
Popular right now
TCL 85-Inch QM6K Series QLED Mini LED 4K HDR Smart Google TV
KitchenAid Artisan 5-Qt Tilt-Head Stand Mixer, Cardamom
Bose QuietComfort Ultra Earbuds (2nd Gen), Deep Plum
As an affiliate, we earn on qualifying purchases.
How Retailers Are Opening Access
Muse’s Amazon block followed a rapid series of announcements. On September 23, Meta said it was adding retail connectors involving Best Buy, Gap, Sephora, Walmart, Wayfair and Expedia. Shopify announced tools in late September that allow browser agents to use structured checkout functions rather than relying only on screen interaction. On October 6 and 9, Sierra and Meta announced the Personal Agent Protocol, also called Poppy, with Walmart, Shopify and Stripe; a draft followed, along with additional design partners.
The approaches fall into three broad patterns. An agent can use a browser much as a person would, an arrangement that can be difficult for a retailer to distinguish from ordinary customer activity. A merchant can provide a dedicated interface, such as Walmart’s announced gateway. Or companies can develop shared protocols to convey an agent’s identity and the customer’s permissions. Poppy was at draft 0.1 in the material provided; other efforts include Google and Shopify’s Universal Commerce Protocol and OpenAI and Stripe’s Agentic Commerce Protocol. The existence of a protocol does not mean all retailers or agents support it.
OpenAI’s experience also shows the gap between building checkout infrastructure and making it work as a product. After launching Instant Checkout in ChatGPT on the Agentic Commerce Protocol in September 2025, OpenAI said in March 2026 that it was letting merchants use their own checkout experiences while the company focused on product discovery. The source material also describes access barriers at other services, suggesting that retailer participation and reliable permissions remain unsettled parts of agent shopping.
“Continued access by an unauthorized AI agent violates Amazon’s Conditions of Use, to which our customers have agreed.”
— Amazon, in a notice shown to Muse users
Access, Safety and Adoption Remain Open
The available information does not establish how many Muse users have completed purchases, how the reported average purchase value was measured or whether it covers only transactions through particular retailers. Meta’s user figures and descriptions of Muse’s capabilities are company statements. Walmart’s announcement also does not specify the gateway’s full availability, participating agents or the practical limits on customer authorization.
Amazon and Meta have made conflicting statements about Muse’s access and handling of credentials. The source material does not provide an independent technical examination resolving those claims. It is also unclear how retailers will distinguish authorized agents from other automated traffic, what happens when a transaction is disputed, or how affiliate referrals and returns will be handled when an agent completes a purchase.
Retailer Integrations Will Test the Model
The next test is whether Walmart’s gateway moves from an announcement to broad use by customers and third-party agents, and whether its spending profiles provide workable controls. Further details about supported agents, purchase approvals and transaction handling would clarify how much of the shopping process Walmart is opening up.
Meta and Amazon may also clarify whether they can agree on access rules and what safeguards would address Amazon’s objections. Meanwhile, protocol developers and retailers will need to show that their systems can work together across discovery, checkout, payment and post-purchase support. For shoppers considering an agent now, the practical questions remain who is authorized to buy, what spending cap applies and which retailer’s return policy governs.
Key Questions
Why did Amazon block Meta’s Muse?
Amazon said Muse was an unauthorized agent, did not identify itself and appeared to capture and store customer credentials. Meta said Muse could not see users’ passwords or payment methods. The source material does not include an independent technical finding resolving the disagreement.
What is Walmart’s Agent Gateway?
Walmart announced a system intended to let approved AI agents search its inventory and make purchases through a dedicated integration, rather than interacting with its website like a person. Walmart also announced Agent Profiles with spending limits.
Are AI agents already completing many online purchases?
The cited September HUMAN Security data found that 81% of agent requests went to product and search pages and 2.4% to checkout and payment pages. That indicates most observed requests were for research, but the figures do not establish the number of completed purchases across online retail.
Could AI shopping agents affect affiliate commissions?
They could, if an agent completes a purchase without using a publisher’s tracked referral link. The source material describes Amazon Associates commissions as tied to tracked clicks, but it does not quantify any resulting loss for publishers.
What should shoppers check before letting an agent buy?
Shoppers should check whether purchases require approval, set a spending cap and understand how the agent handles payment details. They should also confirm which retailer’s return policy applies and how to dispute a purchase.
Source: ThorstenMeyerAI.com
