Meta's consumer AI agent Muse can now place phone calls to businesses in the US, TechCrunch reports. The feature is limited to paid subscribers.
What changed
Muse first tested calling in September. Back then, only a small group of early adopters could use it, and they had to ask the AI for access. Now any paying subscriber can ask Muse to make a call. Alexandr Wang, Meta's chief AI officer, announced the rollout on November 12, 2026. He pitched it as a way to hand off calls people dislike making, such as calls to a cable provider, an insurer, the DMV or a restaurant with no online booking.
Pricing
According to the report, Muse plans start at $20 per month, which includes 500 million Muse tokens per week. Heavier users can pay $100 per month for 3 billion tokens. Calling is paid-only for now. Wang said a separate option, a Muse email address, will soon reach free users.
The article says Muse has more than 10 million users, a figure Meta gave last month.
Not the only one
Muse is not the first agent to phone businesses. TechCrunch notes that the startup Instinct and Google's Gemini offer similar options. The report adds that results vary on the receiving end: many businesses and receptionists end the call once they realize they are talking to an AI. As AI voices get closer to human speech, telling an agent from a customer will become harder.
What this means for the business answering the phone
Until now, most conversation about AI calling has focused on companies placing automated calls. This rollout points the other way: the caller may be a customer's agent, and the person who hangs up may be your own receptionist. The source gives no data on how many calls Muse has placed, how it identifies itself on a call, or whether Meta plans rules for disclosure. Those points are unconfirmed.
What the source does establish is that three products now offer this ability, and that Muse has widened access to everyone on a paid plan. Teams that run phone-based sales or support should treat AI-placed calls as a real category of inbound traffic and set a policy for them.
Top comments (1)
The detail about receptionists hanging up once they realise it's an agent stood out to me. That's the easy case, though.
The harder one is the insurer or cable provider. The call probably comes from a number that doesn't match the customer's file, and it asks about a real account. Staff need a simple rule for what an agent can be told or allowed to change before anyone verifies it.
Would you have teams verify the agent on the call, or call the customer back on the number on file?