Meta Launches Muse: The World’s First Personal AI Agent for Everyday Tasks

Meta has officially introduced Muse, a proactive personal artificial intelligence agent designed to operate 24/7 across connected applications, handle everyday digital tasks, and complete multi-step tasks on behalf of users.

Meta introduced Muse as an independent digital assistant that goes beyond traditional conversational chatbots. Powered by Meta’s Muse Spark large language model, the system is engineered to turn long-term goals into structured action plans.

Muse can connect directly with common personal apps, including email, calendar, payment platforms, health trackers, smart home devices, and shopping sites.

Because of these connections, Muse can handle jobs like booking trips, drafting messages, filling out online forms, organizing calendars, and making purchases on its own.

Security and Virtual Environment

To manage sensitive tasks safely, every user’s copy of Muse runs inside a secure, private cloud-based virtual computer called the Muse Secure VM. This setup allows the software to work on long tasks in the background without needing constant supervision.

Meta added a second safety checker called the Sentinel agent on the same virtual setup, kept completely separate from Muse. Sentinel reviews actions before they happen and stops tasks to ask for user permission before taking important steps, such as sending money. Meta explains that the agent completes tasks without ever seeing raw user passwords or credit card numbers.

Availability and Pricing Tiers

Muse is rolling out first in the United States for users aged 18 and older. Users can access the platform through phone apps on iOS and Android, on the web, and directly inside WhatsApp.

The service comes in three pricing options:

• Free Tier: Offers standard access with a weekly limit of up to 100 million tokens.
• Power Plan: Costs $20 per month for higher usage limits
• Maximum Plan: Costs $100 per month for heavy and continuous task management.