Meta’s artificial intelligence assistant Muse has attracted more than 10 million users in the United States and Canada in approximately six weeks, according to CEO Mark Zuckerberg. The company says shopping has emerged as a leading use case, with more than a quarter of Muse users engaging with its shopping capabilities and the average purchase value exceeding $200. The figures suggest that consumers are beginning to use AI agents for commercial activities rather than relying on them exclusively for answers, research and content generation.
The milestone was announced during Meta’s third-quarter 2026 earnings call as the company outlined its plans to build a larger business around AI-powered personal agents. Around 25% of people who have set up Muse use it on any given day, Zuckerberg said. Meta now wants to improve the service, expand its availability and develop new revenue opportunities. However, the company faces challenges involving computing costs, international expansion, retailer access and consumer trust as AI-driven shopping develops.
Key Takeaways
- More than 10 million users: Muse has attracted over 10 million sign-ups in the United States and Canada in roughly six weeks.
- Shopping is a major use case: More than a quarter of users engage with Muse for shopping.
- Average purchase value: Meta reported an average purchase size above $200.
- Daily engagement: Approximately 25% of people who set up Muse use it on a given day.
- Limited availability: The service has not launched globally because of computing and localization requirements.
- Business opportunity: Meta sees potential in subscriptions, commerce and tools for small businesses.
- Scaling challenge: Improving computing efficiency will be important to expanding the service economically.
Muse Crosses 10 Million Users in Six Weeks
Muse is part of Meta’s effort to develop AI assistants that can help people complete tasks across connected services. Rather than functioning only as a conversational chatbot, the personal agent is designed to work with applications and services to help users manage activities such as shopping, productivity and creative projects.
During Meta’s earnings call, Zuckerberg said more than 10 million people in the United States and Canada had set up Muse within approximately six weeks of its launch. He described early growth and engagement as encouraging, while acknowledging that the company still has considerable work to do.
The geographic limitation is important. Muse has not yet been rolled out globally, partly because the service requires substantial computing resources and must be optimized for different markets and integrations.
The reported milestone also needs context: the figure refers to people who have set up the agent, not necessarily 10 million daily active users. Meta’s separate daily-use figure is approximately 25% of those who set it up.
Why Shopping Has Become a Major Muse Use Case
Shopping is one of the most commercially significant applications for an AI agent because it can connect product discovery with purchasing.
A conventional shopping search often requires consumers to compare products across multiple websites, check specifications, evaluate prices and navigate checkout pages. An AI agent can potentially simplify this process by helping users research options, identify suitable products and complete supported transactions.
Meta says more than a quarter of Muse users engage with shopping and that the average purchase value exceeds $200. The company also said commerce activity exceeded its initial expectations: it worked with payments provider Stripe to raise credit limits during the first week because transaction volumes were higher than anticipated.
These figures suggest that some users are willing to use AI agents for meaningful commercial activity. However, the reported average purchase value does not establish how frequently each user purchases products, how much revenue Meta earns from those transactions or whether the activity is profitable.
How AI shopping agents could change online retail
| Traditional online shopping | AI-assisted shopping |
|---|---|
| Users search across websites themselves | An agent can help research products |
| Consumers compare specifications manually | AI can summarize differences |
| Shoppers navigate supported checkout flows | An agent may help complete a purchase |
| Retailers compete for direct website visits | Agents may become another product-discovery channel |
| Consumers make most decisions themselves | Users can delegate parts of the process |
The degree of automation depends on the retailer, available integrations, permissions and the capabilities of the agent. Not every website or payment flow is supported.
Meta Sees a Larger Opportunity Beyond Shopping
Although shopping is a major use case, Meta says Muse is also used for personal productivity and creative work.
Productivity tasks can include connecting email and calendar services, organizing information and helping users manage their schedules. Creative tasks can involve generating ideas and producing content.
These activities give Meta several possible routes to building a sustainable business around the product.
One option is subscriptions that provide specialized capabilities or expanded usage. Meta has also discussed opportunities to support small businesses that already use Facebook, Instagram and WhatsApp to reach customers.
For businesses, AI agents could help with customer acquisition, routine communications, marketing content and other operational tasks. An agent that helps a customer discover a product and complete a purchase could also create opportunities for merchants to reach buyers through a new channel.
Meta’s challenge is to turn early engagement into recurring usage and a business model that covers the substantial cost of running AI services.
Computing Costs Could Limit Muse’s Expansion
Muse’s early growth does not eliminate the infrastructure challenges associated with scaling AI agents.
Unlike a simple search interface, an agent that reasons through tasks, interacts with connected services and completes multi-step workflows can require significant computing resources. Costs may increase as users assign more complex tasks or use the service more frequently.
Zuckerberg said computing efficiency will be an important constraint over the coming year. Meta wants to improve efficiency while developing enough capacity to support much wider adoption.
The company has expressed ambitions to serve a much larger audience over time, but those ambitions depend on improving the economics of the service. Meta has not disclosed enough information to calculate Muse’s cost per user or its profitability from shopping activity.
International expansion will also require work on local integrations, product optimization and regional requirements. The company must balance growing adoption with ensuring that the service can operate efficiently at scale.
Retailer Restrictions Create Another Challenge
AI-powered shopping could change the relationship between retailers and their customers. If consumers increasingly rely on an external agent to research products and place orders, retailers may have less control over how customers discover products and interact with their brands.
Amazon has already blocked Muse from shopping on its platform, citing security and user-experience concerns, according to reporting by Axios and CBS News.
This illustrates a potential obstacle for agentic commerce: an AI assistant’s usefulness depends partly on access to the stores, product information and checkout systems that consumers want to use.
Retailers may prefer customers to shop directly through their own websites and apps, where they can manage recommendations, collect customer data and maintain the shopping experience. AI companies, meanwhile, want their assistants to become convenient interfaces for product discovery and purchasing.
The result could be a competition over who controls the customer relationship in AI-driven commerce. Wider adoption will depend not only on consumer demand but also on whether retailers and payment providers support these systems.
What the Numbers Mean for Meta’s AI Strategy
Muse’s early adoption provides Meta with evidence that consumers are interested in AI agents that can help with practical tasks. The shopping figures are particularly notable because they suggest that users are willing to move beyond asking questions and use the service in transactions involving real money.
However, sign-ups, daily engagement, completed purchases and profitable monetization are different measures. The next stage will be determining whether users continue returning after the initial launch period and whether Meta can support their activity without excessive infrastructure costs.
Meta also needs to establish reliable integrations, maintain appropriate safeguards and demonstrate that the service provides value to both consumers and merchants. These factors will influence whether Muse develops into a lasting commercial platform rather than a short-lived surge in interest.
The Bigger Picture
Meta’s Muse milestone reflects a broader shift in the AI industry toward agents that can perform tasks rather than merely generate responses. Shopping is an attractive starting point because product research, comparison and checkout involve multiple steps that can potentially be simplified through automation. If the experience proves useful and reliable, AI agents could become another route through which consumers discover products and interact with retailers.
The commercial opportunity remains uncertain. Meta must improve computing efficiency, expand access beyond its initial markets and work through retailer restrictions while building user trust. The company’s early figures indicate demand, but long-term success will depend on repeat usage, transaction reliability, retailer participation and the economics of delivering the service at scale.
FAQs
1. How many users does Meta Muse have?
Meta CEO Mark Zuckerberg said more than 10 million people in the United States and Canada had set up Muse in approximately six weeks.
2. Why is shopping popular on Muse?
Muse can help users research products and interact with supported shopping services. Meta says more than a quarter of its users engage with shopping and that average purchase values exceed $200.
3. Is Muse available worldwide?
No. At the time of the announcement, Muse had not launched globally. Meta cited computing requirements, efficiency and the work needed to support different markets and integrations.
4. How does Meta plan to make money from Muse?
Meta has discussed subscriptions, expanded AI capabilities and opportunities involving commerce and small businesses. The company has not disclosed enough information to determine the agent’s overall profitability.
5. What challenges does Muse face?
The main challenges include computing costs, international expansion, reliable integrations, consumer trust and retailer restrictions. Amazon’s decision to block Muse from shopping on its platform illustrates the difficulty of securing access across the online retail ecosystem.
Looking Ahead
Meta’s next priority will be to improve Muse’s efficiency and reliability while expanding its availability. Continued growth in daily engagement and shopping activity would provide stronger evidence that the agent is becoming a regular part of users’ lives. The company will also need to make its commercial model work without allowing infrastructure costs to grow faster than the value generated by the service.
For online retailers, Muse’s development raises a strategic question: whether AI agents will become an important new source of customers or compete with retailers for control over discovery and transactions. Meta has demonstrated early consumer interest, but the scale and sustainability of AI-driven shopping will become clearer as more merchants participate and the technology expands into additional markets.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
