Meta’s consumer AI agent Muse can now place calls to U.S. businesses on behalf of its users. The feature is available to anyone with a paid Muse subscription, giving subscribers a way to hand off some tedious errands by phone.
Calls move beyond early access
Muse first tested the calling feature in September. At that point, only a select group of early adopters could try it, and they had to ask the AI for access. The feature is now open to all paid subscribers in the U.S.
Meta’s Alexandr Wang, its Chief AI officer and head of its Superintelligence Labs division, described possible uses including calls to a cable company, the DMV, or a restaurant that does not offer online reservations. These are tasks that can take time but may not require the customer to be on the phone themselves.
The change fits Meta’s effort to make consumer AI part of everyday life. The company said last month that Muse had over 10 million users. Taking calls off a person’s to-do list is one way the agent could become useful beyond answering questions.
Calling comes with a subscription
Making calls is currently a paid-only feature. Muse plans start at $20 per month for 500 million Muse tokens per week. Subscribers with heavier AI use can choose a $100 per month plan that includes 3 billion tokens.
That means the feature is tied to a broader subscription, rather than being available to every Muse user. A person who wants help with a business call will need a paid plan, even if the call itself is a simple request such as making a dinner reservation.
Wang said this week that another Muse feature, the option to make a Muse email address, will soon roll out to free users. For now, the difference in access means that some of the agent’s task-handling capabilities are available to a wider audience than its calling feature.
Businesses may not stay on the line
Muse is not the only AI agent that can place calls to businesses. The startup Instinct and Google’s Gemini have offered similar options. The existence of these tools shows that AI agents can take on phone tasks, but a successful call also depends on the business on the other end.
Some businesses and receptionists hang up once they recognize that a caller is an AI rather than a person. That can prevent the agent from completing the task, even if it can manage the conversation. As AI voices become better at mimicking human speech, distinguishing a customer from an AI acting for that customer may become harder.
That raises a practical tension for businesses and callers. An AI agent may be able to wait patiently and follow a phone menu, but businesses may have their own reasons for ending a call when they detect automation. The feature’s usefulness will depend in part on whether companies are willing to interact with it.
A possible answer to frustrating service calls
For consumers, the appeal is straightforward: customer service calls can mean navigating complex phone trees, waiting on hold, and hoping the first representative can resolve the issue rather than transferring the call elsewhere. An AI agent could take on some of that waiting and routine conversation instead.
That does not guarantee a call will be completed or that every company will cooperate. But for people who find these tasks tedious, delegating a call could still be valuable, especially when the request is ordinary and does not require the customer to explain a complicated problem themselves.
Paid subscribers can now ask Muse to make those calls in the U.S. Whether that becomes a dependable way to handle errands will depend not only on the AI’s ability to speak, but also on how businesses respond when an agent calls for a customer.