Muse is quickly becoming a test of whether consumers are ready to hand over more of the shopping journey to an AI agent, from product discovery, price comparisons to the purchase itself. 

In six short weeks, more than 10 million consumers in the U.S. and Canada have signed up for Meta’s AI-powered personal agent, which can proactively perform tasks on a user’s behalf, much like a human personal assistant. It can work in the background to monitor product prices, restock the pantry or plan a trip that fits one’s schedule and budget—with the user giving final approval for a purchase. 

Shopping has emerged as Muse’s most common use, with more than a quarter of weekly users making purchases and transactions averaging over $200, including travel bookings. During Muse’s first week, transaction volume was much higher than expected. As a result, Meta worked with Stripe to raise credit limits to meet demand. 

“Growth continues to be strong and we have been number one in the App Store almost every day,” CEO Mark Zuckerberg said in the third-quarter earnings call. “Engagement and retention are also off to a strong start. About 25% of people who have set up Muse use it on any given day. For our earliest adopters, the longer that they use Muse, the better that retention seems to get.”

Muse’s rapid adoption has drawn a growing roster of retail partners, including Walmart, Best Buy, Sephora, Wayfair, Gap, Ulta Beauty, American Eagle Outfitters, Dick’s Sporting Goods, Fanatics, Michael Kors and eligible Shopify stores—a huge unlock for small direct-to-consumer retailers that run on Shopify. 

For retailers, Muse represents more than another sales channel. It is opening a new gateway for consumers to engage with brands and retailers, much as Meta’s Facebook and Instagram became gateways for product discovery. But unlike social media, which monetized consumer attention through advertising, Muse will generate revenue from the purchases it facilitates, with participating businesses paying a fee for completed transactions. 

And in a potential gamechanger in the agentic AI arms race, Muse is designed to enable a customer’s AI agent to communicate directly with participating retailers’ AI agents to search for products, compare options and facilitate purchases. 

The nation’s largest retailer, Walmart, and a growing list of others have embraced that vision. Amazon, the second largest retailer in the U.S., has not. 

How Walmart Is Managing Muse

To move Muse ahead of Google Gemini and Perplexity, Meta worked directly with business partners, like Walmart, to establish a Personal Agent Protocol so that the shopper’s personal agent could reliably and securely interact with the retailers’ agent while maintaining appropriate controls for both the business and the customer.

Walmart explained why agent-to-agent communication is so vital to the continuing evolution of e-commerce, which was originally designed for how people shop, not agents. 

“Today, agents may shop on Walmart’s site through a web browser, following steps designed for people. This can be slow and error-prone, and agents may get stuck along the way,” it stated. “The way agents shop is different, and so we are building the Walmart Agent Gateway which will give agents a faster, more reliable way to shop.”

To facilitate agentic shopping at Walmart or Sam’s Club, the system will establish a customer-specific Agent Profile, where the customer authorizes that agent to sign into their account and conduct transactions with either a set spending limit or required approval for each transaction. The Agent Profile also stores a customer’s past purchase history within that agent so it can select preferred brands.

Walmart has put a number of safeguards in place to protect the customer. For example, an agent may only take actions that the shopper allows through the Agent Profile. The agent can only access information needed to complete the task at hand and the agent does not receive the customer’s credit-card information stored in Walmart’s system and can’t change their preferred payment method.

Walmart explained: “Put simply, Agent Profiles give customers control over what their agents can do, the Walmart Agent Gateway provides a reliable way for agents to shop and our terms establish the rules that those agents must follow.”

The ultimate success of Muse and its integration with retailers like Walmart depends on building consumer trust. However, Forrester principal analyst Kate Winick observed that overall trust in Meta remains low. “Customers almost expect to have their data mismanaged in exchange for a high level of connectivity and utility,” she said. 

Meta recognizes this vulnerability and has faith it can build that trust over time. Meta CFO Susan Li said, “As trusted integrations develop and agents demonstrate that they can generate incremental value and demand and operate safely, we expect that more and more businesses will be excited to participate.”

Amazon Says No To Meta’s Muse AI Agent

While Walmart welcomes Muse in, Amazon has closed and locked its doors. After asking Meta to voluntarily exclude Amazon from its service, Muse managed to slip in. That forced Amazon to erect an anti-bot wall that prevents Muse from loading the Amazon search page, according to GeekWire. 

Now a user gets a popup stating: “Continued access by an unauthorized AI agent violates Amazon’s Conditions of Use, to which our customer have agreed.”

Amazon disputes Meta’s security and privacy claims, arguing that Muse does not identify itself as an automated agent while browsing and appears to capture and store customer credentials. 

This clash follows efforts by Amazon to block other outside agents, including Perplexity, Google and OpenAI. Amazon views these agents as a threat to its entire commerce ecosystem—bounded by retail sales, advertising, logistics and Prime membership. At stake is the $588 billion in total retail sales Amazon generated last year, including $68 billion in advertising revenue. 

Bain argues that agentic AI threatens to disintermediate the retailer-customer relationship by disrupting the top of the funnel where the shopping journey begins. Deloitte puts it more bluntly: AI agents pose “significant risks to margins, CRM data capture and brand loyalty” by weakening direct customer relationships.

Rather than letting this happen, Amazon is holding tight to its relationship with customers. It’s responded by developing its own agentic shopping capabilities, including Alexa for Shopping across its marketplace and a Buy for Me agent that allows shoppers to purchase products from select third-party brand websites. But unlike Muse, the shopping journey still begins inside Amazon’s ecosystem, maintaining the customer relationship that fuels its e-commerce and advertising business. 

While Walmart is betting that additional e-commerce sales outweighs the risk, Amazon has far more to lose if outside agents become the new gatekeepers between retailers and customers.

Who Wins?

It’s still early days for agentic AI, but the technology is beginning to reshape the entire shopping journey from discovery and research to product selection and purchase. As Bain noted, “While timing and depth of adoption are uncertain, autonomous agents are already redefining the rules of search, advertising, personalization, fulfillment, payments, and post-purchase support. The choices retailers make now will determine whether they shape the future or are sidelined by it.”

Walmart has all to play for by getting in early. The company reported that 23% of its $125 billion in second quarter U.S. sales were generated in e-commerce. Yet unlike Amazon, its competitive moat is the physical, brick-and-mortar store. Walmart’s 4,600 U.S. stores remain the foundation of its customer relationships, giving the retailer reason to open new digital shopping channels to drive additional sales. 

On the other hand, Amazon has primarily a digital presence—Whole Foods is an exception. Its business remains built around digital customer relationships. Opening its platform to third-party shopping agents risks weakening that direct customer connection.

Over the years, Amazon has become the dominant starting point for all online shopping, and it continues to maintain that lead. As of March, Bain reports that 69% of consumers began their online shopping journey with Amazon compared with 53% on Google and 33% on other online websites like Walmart.com. Only 8% started on ChatGPT or other generative AI chatbots. But with the rapid adoption of Muse and other AI agents, consumers shopping behavior could change faster than retailers expect.

Whether Walmart’s openness or Amazon’s defensiveness proves the best approach remains to be seen, as does whether Muse lives up to its early promise.  Meta has a lot at stake. It plans to invest between $135 billion and $145 billion this year, much of it on AI infrastructure and development. 

If Muse succeeds, it could fundamentally reshape Meta’s business model. Today, nearly all company revenue comes from advertising—$63.1 billion of Meta’s $64.6 billion in third quarter revenues were generated by ads. Muse offers Meta a path to a far larger role in e-commerce, beyond feeding ads to consumers to attract their attention and toward benefiting from the purchases that attention inspires.