Meta launches Muse in the United States, but has yet to set a European release date

An agent that performs tasks, not just answers questions

Muse, Meta’s new artificial intelligence application, launched in the United States on September 8, 2026. Unlike a chatbot focused on conversation, it is designed to act on the user’s behalf: it can draft emails, make reservations, fill out forms, and browse websites through a virtual environment in the cloud.

The application can work in the background and, when a service does not offer an interface for connecting systems, interact with the web in a way similar to a person. According to coverage by El Mundo, higher-risk actions—such as sending an email or making a payment—require explicit confirmation. The product description also notes that a separate process, Sentinel, controls which actions are performed and manages sensitive credentials.

Initial adoption in the United States and Canada

Market intelligence firm Apptopia estimated that Muse accumulated 2.8 million mobile installs during its first 12 days. Of these, 1.8 million were on iOS devices in the United States and Canada, according to figures cited by TechCrunch and reported by VnExpress. In those two markets, ChatGPT recorded 1.3 million iOS downloads during the same period following its mobile debut.

The data describe installs and downloads over a specific period: they are not equivalent to active users, subscriptions, or sustained adoption. Even so, Muse rose to the top of the U.S. App Store rankings after its launch, a sign of initial interest in a market where technology companies are competing to turn AI assistants into agents capable of completing tasks.

No date announced for the European Union

The available sources do not report a release date for Muse in Europe announced by Meta. Therefore, neither the launch schedule nor the specific reason for its lack of availability can be treated as confirmed.

A Morgan Stanley note cited by Merca2 points to the obligations that apply to large platforms in Europe. Meta is designated as a gatekeeper under the Digital Markets Act (DMA), and company services such as Facebook, Instagram, and WhatsApp are subject to that regulation. In general terms, the DMA sets conditions for the combined use of personal data across certain services and requires valid consent in applicable cases. The GDPR, in turn, regulates the processing of personal data.

Morgan Stanley’s analysis suggests that these requirements could influence the adaptation and timing of an agent capable of operating across different services. In practice, connecting Muse with other Meta products could require users to expressly select or authorize which services and data are used, according to that interpretation. This is not an explanation confirmed by Meta for Muse’s delay, nor does it mean that the DMA completely prohibits such data combination.

A useful precedent, but not conclusive

Meta AI, a less ambitious product, arrived in Europe in March 2025 after a delay that the company itself then attributed to the difficulties of adapting to the regulatory framework, according to the note cited by Merca2. That precedent shows that European requirements have previously affected the reach and timing of Meta’s AI products, but it does not prove that they are the specific cause of Muse’s situation.

For companies following the development of AI agents, the difference between availability in the United States and Europe matters both in terms of access to the product and the conditions governing data processing and connections between services. For now, the facts confirmed by the sources are the U.S. launch and the absence of an announced European timetable; the role of regulation in that wait remains an analytical hypothesis.