🔍 Read the full analysis: The Retail Battle Over AI Agents And Customer Relationships on ThorstenMeyerAI.com
Get the latest gadgets 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 agent profiles on Oct. 21, while Amazon has blocked Meta’s Muse shopping agent from its site. The clash puts retailers, AI companies and publishers at odds over access, customer relationships and how purchases are credited.
Walmart is opening a dedicated route for approved AI shopping agents, while Amazon has blocked Meta’s Muse from its website, making the two retailers’ opposing approaches a focal point in a dispute over who controls online shopping. The disagreement affects whether agents can compare products and place orders on a shopper’s behalf, and how retailers, AI companies and publishers may earn money from those purchases.
Walmart announced Agent Profiles and a Walmart Agent Gateway on Oct. 21. The gateway uses the Model Context Protocol, or MCP, to let approved agents search inventory and buy through a structured connection rather than navigating the retailer’s consumer website. Agent Profiles are intended to give agents defined permissions, including spending limits. The source material does not provide further implementation details or say which agents can use the gateway now.
Amazon, by contrast, blocked Muse in September. The Verge reported that users trying to use the agent received 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 and argued that the agent did not identify itself and appeared to capture and store customer credentials. Meta’s launch post said Muse does not have visibility into people’s passwords or payment methods. Those accounts describe competing claims; the source material does not independently establish how Muse handled credentials.
Meta said on its Oct. 21 earnings call that more than 10 million people in the United States and Canada had set up Muse within about six weeks, that shopping was its most common use, and that the average purchase exceeded $200. These are company-reported figures, not independently verified usage data. The source material does not specify the period or calculation behind the average purchase figure.
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 policy split matters because an AI agent can sit between a shopper and a retailer. If a user asks an agent to compare prices, choose a product and complete checkout, the agent may influence where the customer shops, what information the retailer receives and which company gets credit for the sale. Retailers want a way to set terms and manage transactions; AI firms want their tools to work across merchants.
There are also consequences for advertising and referral businesses. GeekWire reported that Amazon earned more than $68 billion from advertising last year. Meta chief executive Mark Zuckerberg said the company expected to profit by taking a small fee from transactions. Separately, affiliate publishers may miss commissions if an agent reads their review but completes a purchase without using a tracked link. The source material describes this as a risk, not a measured loss.
For shoppers, convenience comes with practical questions: what the agent is allowed to buy, how spending is capped, and which retailer’s terms apply if an order goes wrong. The source material recommends purchase approval, spending limits and one-time-use cards as safeguards; it does not establish that any particular control is available across all agents or retailers.
Products Worth Considering
Sonos Era 100 Wireless Smart Speaker, Black
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
From Browser Bots to Retail Gateways
Muse launched on Sept. 8 as a personal agent that Meta said could open a browser, fill out forms and negotiate on a user’s behalf. It runs in a dedicated virtual machine in Meta’s cloud and uses Link by Stripe for payment, according to the source material. The browser-based approach can interact with pages designed for people, but retailers may object that the agent is difficult to distinguish from a human user or may not follow site rules.
Meta announced retail connectors with Best Buy, Gap, Sephora, Walmart, Wayfair and Expedia on Sept. 23. Shopify introduced structured tools for browser-agent checkout later that month. In October, Meta and Sierra announced the Personal Agent Protocol, also called Poppy, with Walmart, Shopify and Stripe. A draft followed on Oct. 9 with 35 additional design partners, including OpenAI, Visa, Mastercard, PayPal and Target, according to the source. Poppy was still at draft 0.1, and payments were left to extensions.
Other standards are also in development. Google and Shopify back the Universal Commerce Protocol, while OpenAI and Stripe developed the Agentic Commerce Protocol. Protocols aim to replace fragile interactions with common ways to identify agents, convey customer permissions or handle checkout. But adoption is not settled: in March 2026, OpenAI said its initial Instant Checkout did not offer the flexibility it wanted and that it would let merchants use their own checkout experiences while it focused on product discovery.
“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, as reported by The Verge
Access, Safety and Sales Remain Open
It is not yet clear how many retailers will permit third-party agents to browse or buy, or whether Walmart’s gateway will gain broad adoption. The source material does not establish whether Amazon will change its policy, how disputes over an agent’s credentials and conduct will be resolved, or which company will be responsible when an agent places an incorrect order.
Usage and commercial claims also need careful comparison. Meta’s figures are company-reported, and no independent measure is provided for how often Muse completes purchases or how much revenue it generates. HUMAN Security’s September data, cited in the source material, found that 81% of observed agent requests went to product and search pages and 2.4% to checkout and payment. The figures indicate a gap between research and checkout in that dataset; the source does not provide enough detail here to generalize them to all agents or online shopping.
It remains uncertain whether competing protocols will converge, coexist or be adopted by different retailers. The Poppy draft, the other protocol efforts and the unresolved payment arrangements show that technical connections alone do not settle access rights, consumer protections or how a sale is attributed.
Retailers Set the Next Access Rules
The immediate next step is whether retailers and agent providers turn announced connectors and protocol drafts into functioning, widely available shopping services. Walmart’s gateway will test whether a retailer-controlled connection can offer agents useful access while preserving merchant rules. Amazon’s block will continue to define the limits for agents that rely on ordinary browser interaction, unless the company changes its policy.
Shoppers and businesses should watch for clearer disclosures about agent identity, spending permissions, payment authorization, returns and responsibility for mistakes. Publishers and affiliate businesses will also be looking for mechanisms that credit recommendations when an agent completes the purchase. The source material gives no dates for broader rollout or a resolution of the dispute, so the scale and commercial impact of agent-led shopping remain developing.
Key Questions
What did Walmart announce?
On Oct. 21, Walmart announced Agent Profiles and an Agent Gateway. The gateway uses MCP to let approved agents search inventory and make purchases through a structured connection.
Why did Amazon block Meta’s Muse?
Amazon said Muse was an unauthorized agent, 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 material does not independently resolve the disagreement.
How widely is Muse being used?
Meta reported on Oct. 21 that more than 10 million people in the United States and Canada had set up Muse within about six weeks. The company also said shopping was its most common use; these figures were not independently verified in the source material.
Could AI shopping agents affect affiliate publishers?
Yes, according to the source material. If an agent uses a publisher’s product review but does not follow a tracked affiliate link at checkout, that publisher may not receive commission. The material does not quantify any resulting losses.
What should shoppers check before using an agent?
Shoppers can check what the agent is permitted to buy, whether purchases require approval, how spending limits work, and what payment and return terms apply. The source material recommends limits and one-time-use cards but does not say these options are available with every service.
Source: ThorstenMeyerAI.com
