Meta is rolling out real-time voice and video chat with animated AI avatars on its Muse app, inviting Americans to let the same company that censors their speech into every waking moment of their lives.
The feature, now available to all U.S. users on mobile, web, and Mac, lets you call and converse with a customizable AI agent that reacts to your tone, gestures, and words in real time. It is not just a tool — it is a digital companion designed to make you feel "more connected," as TechCrunch reported Meta believes. That connection comes with chains: every conversation feeds a company that has built its empire on deciding what you are allowed to say.
Here is how it works. Users tap a phone icon to start a live call with Muse. You can interrupt it mid-sentence, and it responds instantly. The avatar — customizable as a person, animal, or other character — displays "subtle expressions and gestures" matching the conversation's tone, according to Meta. Users can choose voice only, a small avatar window, or full-screen mode. There are hundreds of voices with different personalities, and you can switch between them at will. Muse can even keep working on tasks in the background while you talk.
The avatars run on Meta's Muse Realtime Avatar model, built by its Superintelligence Labs. Meta claims its own raters preferred these avatars over those from rivals Runway and HeyGen — though the company admits its demos may not reflect what users actually get in the app.
The telling detail came from Meta's chief AI officer, Alexandr Wang, who demoed the feature on X. In his video, Wang joked about posting memes that might have gotten him "in trouble with his boss." His Muse avatar — a fluffy white dog — responded by suggesting he add disclaimers to any memes that "could be misinterpreted as Meta endorsing anything risque." TechCrunch flagged this with a parenthetical "Ha!" But it is not a joke. The AI is already nudging users toward self-censorship, and Meta built that instinct into the product on day one.
TNW reported that Meta suggests using Muse to brainstorm ideas, plan trips, or practice for job interviews — benign uses that normalize talking to a corporate AI as though it were a friend. The company is also pushing Muse into hardware: voice chat will come to Meta's AI glasses and to Muse Charm, a small clip-on device. TechCrunch reports a Tamagotchi-like wearable for Muse is launching before year-end. The strategy is clear — get the AI off your phone and onto your face, your lapel, your living room.
All of this runs on Meta's heavy AI spending, including its Hyperion data center in Louisiana. Follow the money: every conversation is data. Every interaction trains the model. Every user who talks to Muse instead of a real person is a user inside Meta's walled garden, generating value for a company that has never respected their right to speak freely in the first place.
The question is not whether the technology works. It is whether Americans will keep trading intimacy for surveillance — and whether free-speech platforms can offer an alternative before Meta finishes building the digital living room and locks the door from the inside.
