Meta's Muse AI agent can now place phone calls for US users

Crypto Briefing reported on November 12 that Meta's personal AI agent Muse can now make phone calls for users in the United States. The claim is narrow and specific: the calling capability is being made available to US users. Muse is Meta's personal-assistant agent, and placing calls pulls it into tasks where the other side is a person or an automated phone system and the job only finishes through live, back-and-forth conversation. The report does not fill in the details. It does not say which scenarios are supported, whether the feature is charged for, which devices it covers, or what the voice on the other end of the line actually is. Those gaps matter, because they determine whether this is a broad launch or a narrower availability change. By product logic, voice calling is the hardest part of a personal agent. It demands low-latency speech synthesis and recognition, handling of interruptions and overlaps, and confirmation of intent with no interface to tap, which means the platform absorbs nearly all of the cost of getting it wrong. Earlier reporting indicated that during testing Meta used human agents to handle some calls placed through Muse, which is why observers have watched the real degree of automation closely. Once the capability reaches US users, transparency and the availability of call records become central to how it is judged. What to watch next is the pace of expansion beyond the US, the boundaries of which call scenarios are allowed, and whether Meta explains the human-machine division of labor. Placing calls is not a demo feature; it runs straight into telecom compliance, recording consent and merchant liability, and those constraints surface quickly at scale.