Meta has launched Muse, an AI agent that performs tasks autonomously across apps, including sending emails, filling forms, negotiating, and making purchases. The product is rolling out in the United States on iOS, Android, and muse.ai, but India is not included in the current rollout, Meta confirmed on medianama.com.
Muse operates inside its own virtual machine, with a secondary agent controlling what information it can send to the internet. Meta emphasized that Muse "actually does the work" on behalf of users, such as booking travel or handling emails. The AI agent continues to operate even after the app is closed, distinguishing it from typical chatbots. Muse offers a free tier alongside paid subscriptions at $20 and $100 per month, though Meta has not detailed the differences between these tiers.
This launch is significant because Muse’s architecture is designed with security in mind, a departure from traditional chatbot designs. By running unattended and holding user credentials, Muse raises new considerations for human oversight of agentic AI, a topic currently under regulatory discussion in India. Meta’s approach reflects growing industry attention to the risks and controls needed for AI agents that act autonomously on users’ behalf.
Meta’s Muse is currently available only in the United States, with no announced timeline for expansion to other markets, including India. The product’s rollout on iOS, Android, and muse.ai marks a notable step in AI agent capabilities and user interaction models, according to medianama.com.