Meta’s Muse Gains Millions of Users as Retailers Report Little Disruption and Amazon Blocks Access

Meta’s Muse has attracted more than 10 million users in the United States and Canada since its September release, but its commercial impact remains limited. More than 25% of users have tried the assistant for shopping, and Meta says average purchases exceed $200. The agent connects with calendars, email, and phone functions to complete tasks on a user’s behalf, positioning shopping as its leading consumer use case.

Muse’s early popularity has sharpened concerns about AI agents changing how people discover products, compare prices, and maintain relationships with brands. The app briefly overtook ChatGPT as the top free app on Apple’s App Store, while Meta signed partnerships with Walmart, Best Buy, Gap, Shopify, and other retailers. The adoption figures show reach, but retailer transaction data indicates that reach has yet to translate into a broad commercial shift.

Companies across travel, telecommunications, food delivery, banking, insurance, fitness, and storage report that agent-driven activity is still small. DoorDash counted orders from Muse and other agents in the hundreds. T-Mobile said activity had not reached a significant level, while Hilton described bookings as insignificant. Ally Financial, eBay, Life Time Group, Public Storage, Hartford Insurance Group, and India’s Eternal also reported very low or zero measurable effects.

Those responses contrast with Wall Street’s concern that proactive shopping agents will weaken brand loyalty, intensify price competition, cancel unused subscriptions, and pressure margins. Executives are generally presenting agents as an opportunity, in part because adoption remains too early to threaten established customer channels. Many companies have only recently begun tracking agent activity, and several retail partners have not yet disclosed post-launch results.

Amazon blocked Muse from accessing its online store in September, saying the assistant’s access violated customer-agreed terms of use. The decision reflects a direct conflict between Meta’s agent-led shopping model and Amazon’s controlled retail funnel, where website visits expose shoppers to advertising and Amazon’s own merchandising systems.

Amazon CEO Andy Jassy told analysts that outside shopping agents are still “not great” and no more convenient than visiting a retailer directly. He also said such agents did not cause Amazon’s weaker-than-expected third-quarter advertising performance. Amazon is promoting its own Alexa shopping agent instead, giving the company control over customer interactions and purchase data. Jassy said more than 400 million customers used Alexa for shopping during the past year, with interactions up 570% year over year and US users spending 50% more per order than nonusers.

The central obstacle for Muse is execution. Consumers may want an assistant to search, compare, and purchase products, but the system has to handle discounts, loyalty programs, authentication, inventory, and checkout without creating extra work. Chipotle’s digital team found that agents struggled with loyalty discounts and ordering problems, an early sign that ordinary web infrastructure is poorly suited to autonomous purchasing.

That friction helps explain why high user counts have produced few visible changes in corporate results. A person can activate Muse, ask it to shop, and still abandon the process when an agent hits a retailer-specific rule. Amazon’s criticism focuses on the same weakness from another angle: the agent has not yet made buying easier than using a familiar store.

Companies are already adjusting their websites and commercial strategies despite limited present-day volume. Retailers that partner with Meta are testing ways to make catalogs, pricing, promotions, and checkout accessible to Muse, while other businesses are monitoring whether agents generate cancellations, bookings, support requests, or purchases.

The split between partners and blockers will shape the next stage of agentic commerce. Walmart, Best Buy, Gap, and Shopify have accepted arrangements that give Muse a route into shopping, while Amazon is defending its own interface and advertising model. Other businesses face a choice between opening systems to outside agents, building proprietary assistants, or waiting for clearer demand.

Muse has established consumer awareness faster than it has established a new buying habit. Meta’s user and purchase figures indicate strong curiosity, while retailer disclosures show that autonomous commerce remains a small share of activity. The gap reflects a product that has reach but still encounters practical barriers across the checkout stack.

The next evidence will come from retailers that partnered with Meta and from companies that begin reporting agent-originated transactions separately. Amazon’s Alexa figures also provide a competing model, centered on an assistant controlled by the retailer rather than an outside platform. If Muse becomes more reliable and convenient, businesses will face pressure to support it. Until then, executives are preparing for disruption that has yet to appear in their numbers.