🔍 Read the full analysis: AI Buys, Retailers Adapt: Amazon And Walmart Take Different Routes 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 has announced an Agent Gateway and per-agent profiles that let approved AI agents search products and make purchases under set permissions. Amazon has blocked Meta’s Muse agent, exposing a divide over access, customer relationships and control of online checkout. Adoption figures and agent traffic data cited here come from company statements and a third-party report, and do not establish how often agents complete purchases.
Walmart is opening a route for approved AI shopping agents, while Amazon has blocked Meta’s Muse, setting two major retailers on different paths as software agents begin shopping for consumers. Walmart announced its Agent Gateway and agent-specific profiles on October 21; the dispute raises practical questions about which products agents can compare, how purchases are approved and which company controls the customer relationship.
Walmart’s Agent Gateway is designed to let approved agents search inventory and buy through a structured connection rather than interact with the retailer’s website as if they were human users. Walmart also announced Agent Profiles, which can set permissions and spending limits for an agent. The announcement describes the planned access model; details on availability, participating agents and the full range of controls were not provided in the source material.
Amazon, by contrast, has blocked Meta’s Muse. Users reportedly encountered a notice saying continued access by an unauthorized AI agent violated Amazon’s Conditions of Use. Amazon said Muse had not identified itself as an agent and raised concerns about how it handled customer credentials. Meta’s launch post disputed the password concern, saying Muse had no visibility into users’ passwords or payment methods. Those are opposing company accounts; the available material does not independently resolve the dispute.
Meta said during 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, and that shopping was its most common use. Meta also reported an average purchase above $200. These are Meta’s own figures, not independently verified transaction data. A September report from HUMAN Security, as summarized in the source, found that 81% of agent requests went to product and search pages, while 2.4% went to checkout and payment. That split suggests agents may be used more for research than completed purchases, but it does not establish the overall share of retail sales made by agents.
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 choice between an open agent connection and a blocked one could shape where consumers can comparison-shop and how easily an AI assistant can complete a transaction. A retailer-approved gateway can give a store more control over what an agent sees and what actions it may take. Browser-based agents can attempt to use ordinary shopping pages, but retailers may object that they are difficult to distinguish from people or may interact with accounts in ways the retailer has not authorized.
The commercial stakes extend beyond the sale itself. Retailers rely on advertising, product placement and direct customer relationships; publishers and affiliate sites may rely on tracked clicks to earn commissions. If an agent reads a review but makes a purchase without using an affiliate link, the publisher may not receive credit under arrangements that require a tracked click. The source says Amazon earned more than $68 billion from advertising last year, citing GeekWire, and reports Meta’s stated aim to take a small fee from transactions. Those figures and intentions point to competing business models, not proof of how revenue will ultimately be divided.
For shoppers, the immediate concern is control. An agent that can act on a person’s behalf may make buying easier, but users need to know what it is permitted to buy, how much it can spend and which seller’s return rules apply. Walmart’s announced profiles address some permission questions, though their operation and coverage remain to be seen.
Products Worth Considering
Sonos Era 100 Wireless Smart Speaker, Black
Bose SoundLink Flex Bluetooth Speaker (2nd Gen), Portable Waterproof Speaker, Black
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
As an affiliate, we earn on qualifying purchases.
From Browser Bots to Retailer Gateways
The current split follows a rapid series of product and protocol announcements. Meta introduced Muse on September 8, describing it as an agent that can browse, fill out forms and negotiate for a user. The source says Muse runs in a dedicated computer in Meta’s cloud and uses Link by Stripe for payments, including a one-time-use card. On September 20, Amazon blocked the agent. Three days later, Meta announced retail connectors with Walmart and several other companies, including Best Buy, Gap, Sephora, Wayfair and Expedia.
Other companies are testing ways to make agent commerce more structured. Shopify introduced tools for browser agents to use during checkout, while Meta, Sierra and Walmart have worked on the Personal Agent Protocol, also called Poppy. A draft published October 9 describes a way to identify an agent and record what a customer has authorized; the source says payment functions were left to extensions. Google and Shopify’s Universal Commerce Protocol addresses more of the shopping process, while OpenAI and Stripe’s Agentic Commerce Protocol focuses on checkout. These efforts are at different stages and should not be treated as a single settled standard.
The history also shows that technical integration does not guarantee widespread agent checkout. OpenAI launched Instant Checkout in ChatGPT in September 2025 using its protocol with Stripe, then said in March 2026 that it was stepping back from the initial checkout approach to focus on product discovery and let merchants use their own checkout experiences. That change is a reminder that discovery, payment and merchant control can be difficult to combine in one system.
“Muse has no visibility into people’s passwords or payment methods.”
— Meta, in its Muse launch post
Agent Access and Purchase Rules Remain Open
Several practical details are not established in the source material. It does not say when Walmart’s gateway will be broadly available, which agents will be approved, or how its profiles will work across different retailers. It is also unclear how Amazon will treat other agents, whether its access restrictions will change, or how the companies might resolve their disagreement over Muse’s identity and credentials.
The reported adoption and traffic figures need careful interpretation. Meta’s setup count and average purchase value are company-reported metrics; the source gives no independent audit or breakdown showing how many users completed purchases through Muse. HUMAN Security’s request percentages describe the traffic it measured, but the supplied material does not provide the full methodology or population covered. They should not be read as a market-wide measure of agent shopping.
It is also unclear how customer protections, refunds, returns and affiliate commissions will work when an agent selects a product or completes a purchase. Competing protocols may reduce technical barriers, but their adoption and terms are still developing. The source’s descriptions do not establish that the different protocols are interoperable or that a retailer will accept every agent using them.
Watch Gateway Rollout and Agent Rules
The next developments to watch are Walmart’s rollout details: when its Agent Gateway opens, which agent providers can connect and what spending and purchase approvals customers can set. Walmart’s partners and other retailers may also clarify how agents will identify themselves and whether a purchase requires confirmation before checkout.
On Amazon’s side, further statements could clarify whether its restriction applies specifically to Muse or reflects a broader policy for outside agents. The protocol groups are expected to continue developing their specifications, but the source material does not give a confirmed timeline for full deployment. Until those rules and integrations are clearer, consumers should check an agent’s permissions, purchase limits, payment method and the retailer’s return policy before allowing it to buy.
Key Questions
What did Walmart announce?
Walmart announced an Agent Gateway for approved AI agents to search inventory and make purchases through a structured connection. It also announced Agent Profiles intended to set permissions and spending limits. The supplied source does not specify a broad launch date or list of supported agents.
Why did Amazon block Meta’s Muse?
Amazon said Muse was an unauthorized agent that had not identified itself and raised concerns about customer credentials. Meta said Muse had no visibility into users’ passwords or payment methods. The source reports the companies’ opposing positions but does not independently settle the dispute.
Are AI agents already completing many online purchases?
The source cites HUMAN Security data showing that, within the measured agent requests, 81% went to product and search pages and 2.4% to checkout and payment. That suggests more activity was focused on research than checkout in that dataset. It does not establish what share of all online purchases is made by agents.
How could agent shopping affect affiliate publishers?
Affiliate commissions often depend on a shopper using a tracked link. If an agent reads a publisher’s review but later buys directly from a retailer or through another route, the publisher may not receive a commission under that arrangement. The scale of this effect is not established in the source.
What should shoppers check before letting an agent buy?
Users should review purchase approval settings, spending limits, payment details and return rules. They should also understand which retailer is responsible for the order and how to dispute or return a purchase made by an agent.
Source: ThorstenMeyerAI.com
