MENLO PARK, Calif. — Meta Platforms is betting that Muse, its new artificial intelligence personal agent, can turn early app store success into a lasting consumer habit. It could even establish a new source of growth beyond social media advertising. The app has quickly climbed download rankings, drawing investor enthusiasm, but questions about privacy, reliability and profitability remain as the company tries to persuade users to let AI handle more of their daily lives.

Meta CEO Mark Zuckerberg has described Muse as a centerpiece of the company's artificial intelligence strategy. Unlike conventional chatbots that primarily answer questions, Muse is designed to carry out tasks, including organizing information, researching purchases, drafting emails and helping users plan projects.

The challenge now is to demonstrate that consumers will continue using the service after the novelty fades. For Meta, which owns Facebook, Instagram and WhatsApp, the success of Muse could help determine whether personal AI agents become a major new technology business or an expensive experiment.

Muse Moves Beyond the Traditional Chatbot

Meta introduced Muse on Sept. 8, 2026, initially making it available to adults in the United States through a dedicated application and WhatsApp. The company says the agent is powered by Muse Spark, its model designed for tasks requiring planning and action rather than simple question-and-answer exchanges.

Muse operates through a conversational interface that allows users to describe a goal and delegate parts of the work. Depending on the task and permissions granted, it can research information, help with shopping, prepare communications and work across connected services.

Meta says Muse runs through a dedicated virtual machine, called Muse Secure VM, that provides the agent with its own browser and a controlled environment for handling tasks and user data. The company says users determine how much access the assistant receives.

The service also is expanding beyond smartphones. Meta has announced plans for additional desktop capabilities, video interactions with customizable avatars, email functionality and integrations with other applications. The company is exploring ways to connect Muse with its smart glasses, potentially allowing users to interact with the assistant while moving through the physical world.

Those features could make Muse more useful than a conventional chatbot, but they also increase the complexity of protecting personal information and preventing unintended actions.

Downloads Generate Investor Enthusiasm

Muse's early reception has attracted attention from investors and technology analysts. Data from market intelligence firm Sensor Tower showed the app climbing to the top of U.S. app store rankings shortly after its launch. Subsequent estimates from Sensor Tower indicated that Muse had surpassed 3.4 million downloads by late September, although figures from different analytics firms have varied.

The company's broader strategy is important to investors because Meta remains heavily dependent on advertising revenue generated by its social platforms.

Facebook and Instagram provide Meta with enormous audiences and extensive advertising infrastructure. Muse offers a different potential business: an AI assistant that consumers might use to accomplish tasks rather than simply view content.

A successful agent could deepen Meta's relationships with users, create opportunities for paid subscriptions and help businesses connect with customers. It could also strengthen the company's wider AI ecosystem by encouraging users to adopt Meta's models and services across multiple devices.

However, downloads do not necessarily translate into sustained engagement, paying customers or profits. The more meaningful measures will include how often users return, which tasks they delegate, how reliably the agent completes them and whether customers are willing to pay for additional capabilities.

Subscription Plans Raise Questions About Profitability

Meta is offering a free tier of Muse alongside paid subscription plans reported at $20 and $100 per month, with access levels tied to tokens, a measure used to calculate AI processing requirements.

The free option lowers the barrier to adoption and allows Meta to learn which features users value most. Paid plans could eventually provide a direct revenue stream, particularly for people who rely on the assistant for complex or frequent tasks.

But AI agents can be expensive to operate. Every request may require computing resources, and tasks involving multiple websites, extended reasoning or repeated interactions can consume substantially more processing power than a simple chatbot response.

Meta must determine whether subscription income, commercial partnerships or other revenue sources can cover those costs. A generous free tier could accelerate adoption while also increasing expenses if users make heavy demands on the service.

The company also faces a challenge in applying its traditional advertising model. Consumers may be reluctant to share sensitive information with an assistant that could access personal communications, shopping habits and schedules. That hesitation could limit the use of data for targeted advertising, even if such advertising eventually becomes part of the business strategy.

Trust and Privacy Remain Central Concerns

For Meta, the central question is whether consumers will trust the company with a more intimate view of their lives.

Meredith Whalen, chief research officer at IDC, said in an email that Muse could become a frequently used service because it completes tasks rather than simply responding to questions.

“The real question is trust,” Whalen said.

The concern is particularly significant given Meta's history of privacy controversies and scrutiny of its social media platforms. The company must persuade users that an agent with access to personal information will operate transparently and respect the limits they establish.

Meta has emphasized the security architecture behind Muse, but independent scrutiny and users' experiences will help determine whether those assurances are sufficient.

The stakes are not limited to data security. An agent that sends an incorrect message, makes an unauthorized purchase or mishandles a booking could undermine confidence quickly. Users also need to understand when an AI system is acting autonomously, what information it has accessed and how to stop or reverse an action.

Dazzle and Other Startups Pursue the Same Opportunity

Meta is not alone in trying to make personal AI assistants part of everyday life.

Dazzle, a startup founded by former Yahoo CEO Marissa Mayer, is using a free-to-start approach to study consumer behavior and identify features that people consider valuable. Its photo-focused assistant analyzes personal images to learn about users' interests and relationships, then offers recommendations and assistance with tasks such as finding deals.

Mayer has said that early usage patterns will help guide product development and future pricing. Dazzle is considering subscriptions and transaction fees as possible business models.

Wajo AI is also exploring consumer demand for digital assistants that can research information, arrange appointments and order gifts. Its founder, Shivani Poddar, has suggested that customers might be more willing to pay for completed outcomes than for a conventional subscription.

These companies face a common challenge: identifying tasks that are valuable enough for consumers to delegate regularly and pay to have completed.

Meta Has the Resources to Wait

Meta has a substantial advantage over many startups because its advertising business generates the resources needed to invest in AI infrastructure and subsidize consumer adoption.

That financial strength gives the company time to experiment with pricing and features. However, it does not eliminate the risk that operating costs will outpace revenue or that consumers will prefer competing assistants.

For Zuckerberg, Muse represents a broader effort to position Meta for a future in which AI systems mediate more online activity. If consumers begin asking assistants to compare prices, arrange services and complete purchases, the companies controlling those assistants could gain influence over how people interact with businesses.

The outcome remains uncertain. Muse's initial popularity demonstrates consumer curiosity, but long-term adoption will depend on usefulness, reliability, privacy and cost.

Meta has built an app that is attracting attention. Its next task is to prove that people will trust it with meaningful responsibilities — and that the resulting business can be profitable.