Meta has launched Muse, a new personal AI agent designed to do more than answer questions. It can work across connected apps, send emails, book travel, fill out forms, make purchases and continue working in the background after a user closes the app.
The launch marks a major step toward consumer AI agents that can take actions on a person’s behalf rather than simply generate text.
What happened
Meta announced Muse on September 8, 2026. The company describes it as a personal AI agent built for everyday use, powered by Muse Spark, Meta’s most capable model for agentic work.
Muse can open its own browser, complete forms, negotiate on a user’s behalf and keep working on longer tasks. It can also return for approval when a sensitive action is required, such as sending an email or completing a purchase.
Meta says Muse is rolling out in the United States on iOS, Android and the web, with support for AI glasses coming later.
What Muse can actually do
Examples highlighted by Meta include sending emails, booking travel, selling a car, lowering a bill, organising schedules, turning an Instagram recipe into a grocery list and remembering personal preferences for future tasks.
For purchases, Muse can use Link by Stripe. Meta says Link can generate a one-time-use card so a user’s real card details remain hidden. Shop Pay and 1Password support are also planned.
How Meta is handling security
Because an AI agent can act across personal accounts, Meta has built Muse around a dedicated cloud computer called Muse Secure VM. Each user’s agent runs in its own isolated virtual machine, where connected-service credentials and data are stored.
A separate Sentinel agent supervises what Muse does. Meta says nothing Muse sends to the internet is allowed through unless Sentinel approves it, and users are asked for permission before sensitive actions.
Meta also says Muse cannot directly see passwords or payment details, connected services can be revoked at any time, and users can review an audit trail of what the agent has done and plans to do.
Why this matters
AI assistants are rapidly shifting from systems that answer questions to systems that can carry out multi-step tasks. Muse is especially notable because Meta can distribute it through consumer products such as WhatsApp and potentially its AI glasses.
That could give agentic AI a path to a much broader audience than specialist tools aimed primarily at developers or enterprise users.
At the same time, the more capable these systems become, the more important security, permissions and reliability become. Reuters reported that internal testing raised concerns including disconnections, unprompted data uploads and security issues. Meta says Muse meets its safety and privacy standards but acknowledges that AI agents can still make mistakes.
What it means for readers
For now, Muse is initially rolling out in the United States, so availability is limited. But the launch is worth watching because it points toward a future where an AI assistant does not just tell you how to complete a task — it completes much of the task for you.
If Muse expands internationally and into WhatsApp and AI glasses at scale, Meta could become one of the biggest consumer distribution channels for autonomous AI agents.
Sources
Meta — Introducing Muse: The World’s First Personal AI Agent Built for Everyone
Reuters — Meta launches AI agent that can access other apps to send emails, make payments
