Meta has launched Muse, a personal AI agent designed to carry out tasks across apps and online services on behalf of users, marking the company's latest move beyond conversational chatbots towards AI systems capable of taking actions.
Muse is initially rolling out in the United States for users aged 18 and above. It can be accessed through a dedicated app, the web and WhatsApp, with Meta also planning to bring the technology to its AI glasses.
Powered by Meta's Muse Spark model, the agent can connect to services including email, calendars, payments, shopping, dining, health and fitness applications, smart home platforms and other digital services. Users can choose which services they want to connect and determine the level of access provided.
Muse can perform tasks such as drafting and sending emails, booking travel, filling out online forms, creating shopping lists, sending invitations and making purchases. Meta also says the agent can work on longer tasks in the background after a user closes the app, returning when it requires additional information or approval.
For example, users could ask Muse to plan a trip, compare available options and help complete bookings. The system can also turn a saved recipe into a grocery list or assist with tasks such as negotiating bills or prices.
Purchases are supported through Stripe's Link checkout service, while additional integrations are expected. Meta says users will be asked for approval before certain sensitive actions are completed.
The broader access required by an autonomous AI agent also raises privacy and security considerations. Muse operates inside what Meta calls a Secure VM, a dedicated cloud-based virtual machine intended to isolate the agent and its connected information.
A separate system called Sentinel monitors actions leaving the virtual environment and can request user approval when necessary. Meta says users can disconnect services or change permissions, while an activity trail allows them to review actions taken by the agent.
Muse was previously known internally as Hatch and forms part of Meta CEO Mark Zuckerberg's wider push towards what the company describes as personal superintelligence.
The launch puts Meta more directly into the growing market for agentic AI, where technology companies are developing systems that can move beyond generating responses and execute multi-step digital tasks.
For Meta, Muse will test whether consumers are willing to give AI systems greater access to everyday digital services in exchange for more autonomous assistance.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.