Meta’s Muse AI Agent Surpasses 10 Million Users as Shopping Drives Adoption

Meta Platforms' personal artificial intelligence agent, Muse, has attracted more than 10 million users since its September launch, with shopping emerging as its most popular use case and offering the company an early route to monetizing its expanding AI investments.

Chief Executive Mark Zuckerberg disclosed the figures on Wednesday during Meta's third-quarter earnings call, saying user growth "continues to be strong". More than a quarter of Muse users have made purchases through the application, with the average transaction exceeding $200, he said.

The figures suggest that Meta is gaining early traction with an AI product designed to move beyond answering questions and generating content towards completing tasks and facilitating transactions. If that behavior persists, Muse is expected to give the company another way to generate revenue from its vast consumer audience, extending its business beyond advertising.

Meta shares rose 5% in after-hours trading on Wednesday following the earnings announcement.

Muse, which Meta says can assist with tasks such as sending emails and booking travel, climbed to the top of Apple's App Store rankings shortly after its release. Its early adoption provides a potential commercial foothold for Zuckerberg as Meta commits substantial resources to developing AI models, infrastructure and consumer-facing products.

At the company's annual Connect event, Zuckerberg said Meta expected to monetize Muse by "taking a small fee from transactions". The approach would allow the company to earn revenue when users purchase goods or services through the agent, rather than relying exclusively on advertising or subscriptions.

The shopping figures are significant because they suggest users are already turning to Muse for activities with a direct commercial outcome. An AI agent that helps consumers discover products and complete purchases could become an intermediary between shoppers and merchants, potentially giving Meta a role in transactions that currently take place across retail websites and applications.

The reported average purchase size of more than $200 also points to transactions with meaningful value. However, Meta has not disclosed the total value of purchases made through Muse, the frequency of repeat purchases, or how much revenue the application has generated so far.

Those details will be important in assessing whether the initial user numbers can translate into a durable business. More than 10 million users marks an encouraging starting point, but sustained engagement, repeat transactions and the cost of operating the service will determine its commercial significance.

The proposed transaction-fee model also introduces questions about how Meta will balance monetization with the user experience. The company will need to make the purchasing process sufficiently useful to encourage adoption while ensuring that commercial incentives do not undermine confidence in the agent's recommendations.

If Muse becomes a regular starting point for shopping, Meta could gain influence over product discovery and purchasing decisions. That would create opportunities to earn fees from transactions and potentially strengthen its relationships with merchants.

Despite the early growth, consumer trust could prove a major constraint on Muse's expansion, particularly because personal AI agents may need access to information about users' preferences, communications and plans to perform tasks effectively.

Kate Winick, a principal analyst at Forrester, said consumer trust was the biggest uncertainty in Meta's AI ambitions.

"The more information Muse has, the more effective it can be, and overall trust in Meta remains low," Winick said.

She argued that Meta has developed a reputation that has not prevented continued engagement despite concerns about how it handles user data.

"But Meta has cultivated a reputation that negates trust as a condition for engagement: consumers almost expect to have their data mismanaged in exchange for a high level of connectivity and utility, and have shown that they're willing to keep showing up as long as the product experience delivers," she added.

Users of AI agents may tolerate advertising and data collection on social media platforms while remaining more cautious about allowing an automated assistant to access personal information, make recommendations or carry out purchases on their behalf.

The more tasks an agent can perform, the greater its potential utility, but the more important transparency, privacy safeguards, and user control become. A single poor recommendation or mishandled transaction could damage confidence in a product whose value depends on users delegating decisions to it.

Meta's legal and regulatory challenges add another dimension to that risk. On Wednesday, a New Mexico judge ordered the company to pay $25 billion over allegations concerning misleading users about its handling of their data and its failure to prevent misinformation on Facebook.

The ruling could intensify scrutiny of the company's data practices at a time when it is seeking to persuade consumers to use a more personalized form of AI. The case also highlights the tension between Meta's ability to build highly engaging services using consumer data and the legal and reputational consequences associated with how that data is managed.

Muse's early performance gives Meta evidence that consumers are willing to experiment with AI agents, particularly when they offer practical benefits such as shopping assistance. Yet the number of users and the value of transactions are only the first indicators of a business whose economics remain unclear.

Now, Meta must prove that users return regularly, that transaction fees can generate meaningful revenue, and that the cost of developing and operating the service does not outweigh its contribution. It will also need to establish how Muse can compete with other AI assistants and shopping platforms seeking to occupy the same position in consumers' digital lives.

The wider opportunity is to turn AI from a feature that attracts attention into a service that facilitates measurable economic activity. Shopping provides an identifiable route towards that goal because purchases create a direct link between usage and potential revenue.