Meta’s AI agent, Muse, has officially reached a new milestone in its evolution toward becoming a truly autonomous personal assistant. In a significant update to its functionality, the platform now allows users to delegate time-consuming phone calls to the AI, provided they are enrolled in a paid subscription plan. This shift represents a broader push by Meta to transform Muse from a passive conversational tool into an active agent capable of interacting with the physical world on behalf of its users.
The new capability enables Muse to place calls to businesses to handle a variety of administrative tasks, such as making restaurant reservations, scheduling medical or service appointments, and managing other logistical errands that users typically find tedious or time-consuming. According to Alexandr Wang, Meta’s Chief AI officer, the feature is designed to handle the "calls you don’t want to make," effectively offloading the friction of modern bureaucracy to an automated system.
This functionality is currently gated behind a premium tier of the Muse service. Subscribers who pay for this higher level of access receive not only these advanced calling capabilities but also a larger allocation of tokens and access to specialized features, positioning the paid version of Muse as a robust productivity suite rather than just a casual chat interface.
Early adopters and internal testers have already begun putting the technology through its paces. Reports from those with early access suggest that Muse is proving to be surprisingly adept at navigating the nuances of human-to-human phone interaction. Among those sharing their experiences is Matt Schlicht, a Moltbook creator and Meta Superintelligence employee, who recently provided a practical demonstration of the agent’s utility. Schlicht shared a recording of an interaction where Muse contacted a Volvo dealership to inquire about the next available service appointment.
The recording highlighted a key advantage of AI-driven calling: patience. In the conversation, the Muse agent successfully waited on the line for 15 minutes while navigating the dealership’s hold system. This ability to endure long wait times—often the most frustrating part of calling service departments or support centers—is a primary value proposition for the service. Schlicht noted that he now utilizes the agent for multiple calls of this nature on a daily basis as he continues to refine his approach to integrating the tool into his personal workflow.
Despite the apparent success of these early tests, the implementation of AI-driven phone calls raises interesting questions about transparency and etiquette in communication. In the recording shared by Schlicht, the Muse agent does not explicitly state that it is an artificial intelligence. Instead, it identifies itself as calling on behalf of the user and mentions that the call is being transcribed. While this may be sufficient for accomplishing straightforward tasks, it introduces a layer of ambiguity. There is a persistent risk that such interactions could encounter friction, particularly if the person on the receiving end becomes confused or decides to terminate the call upon realizing they are speaking with a machine rather than a human. As these technologies become more prevalent, businesses may need to develop new policies regarding how they handle calls from autonomous agents.
Meta’s rapid expansion of Muse’s feature set underscores the company’s aggressive strategy in the artificial intelligence sector. Since its introduction, the platform has gained significant traction, surpassing 10 million users. To keep this momentum, Meta has been consistently rolling out integrations designed to make the agent more useful in a professional and personal capacity. One such recent addition is the ability for users to assign a dedicated email address to their Muse agent. This allows the AI to be looped into important email threads and communications without the user having to grant the system wholesale, unmonitored access to their entire personal or professional inbox. This approach balances the need for utility with data privacy concerns, giving users more granular control over how much information their AI assistant can see and process.
Meta is by no means the only major technology player exploring this territory. The competition to create the "ultimate" personal assistant has become a focal point for several Silicon Valley giants. Google, for instance, has been iterating on its own suite of AI calling features for years. Most notably, the company introduced "Call For Me" for its Pixel smartphone line, a feature specifically designed to handle the wait times and navigation of automated phone menus on behalf of the user. Similarly, the startup Instinct has recently debuted a "Concierge" feature for its own AI agent, which is specifically optimized to manage more complex, multi-step calls that require a higher level of reasoning and context.
These developments signal a shift in how the public interacts with generative AI. We are moving away from an era where AI is primarily used for generating text or images toward an era where AI is expected to interface with existing human systems. By enabling Muse to place phone calls, Meta is essentially bridging the gap between digital intelligence and the legacy infrastructure of the telephone. While the technology is clearly in its growth phase, the ability to outsource the "drudgery" of daily life—waiting on hold, checking appointment availability, and managing basic scheduling—is likely to become a standard expectation for premium AI services.
As Meta continues to refine Muse, the company faces the dual challenge of improving the agent’s conversational naturalness and ensuring that its autonomous actions remain reliable and well-received by the businesses being called. The success of this feature will likely depend on how well the AI handles the unpredictable nature of human conversation, including complex hold systems, transferred calls, and the reactions of the individuals on the other end of the line. For now, however, the ability to reclaim 15 minutes of one’s day by letting an AI handle a service call is a tangible, high-value utility that is already resonating with the platform’s most dedicated users.
