Meta launches Muse with a dedicated cloud computer

Meta introduced Muse on 8 September as a personal AI agent running on a dedicated cloud virtual machine. The launch announcement describes an agent that can work across connected services, with users communicating through its own app or WhatsApp.

Meta says a separate Sentinel agent checks outbound activity, and Muse asks for approval before sensitive actions such as sending email or making purchases. People can choose the services they connect, change permissions and review an activity trail.

Meta says Muse conversations and virtual-machine data do not feed its advertising systems. Users can also opt out of interactions being used to train AI models. A Confidential VM, encrypted with a key held only by the user, was described as a later release. Initial distribution was in the United States on iOS, Android and the web. The launch therefore combines a persistent agent with specific controls, while some privacy technology remained a plan.