Muse can now place calls to US businesses — if you pay

Meta announced that Muse, its consumer‑focused generative AI, now dials U.S. business lines on behalf of users, but the capability is locked behind the $20‑per‑month (500 million token) and $100‑per‑month (3 billion token) subscription tiers. The rollout follows a limited beta in September and a public tweet from chief AI officer Alexandr Wang, who framed the feature as a way to offload “annoying” calls to cable firms, DMVs, or restaurants. By converting text prompts into spoken requests, Muse aims to embed itself deeper into everyday workflows and justify its token‑based pricing model.

The move pits Meta against niche players like Instinct and tech giants such as Google, whose Gemini model already supports outbound calls. While AI‑generated speech has reached a level where receptionists often cannot tell the difference, many businesses still terminate calls once they suspect a non‑human caller. Muse’s entry underscores a broader industry push to turn conversational agents into multi‑modal assistants that can navigate phone trees, a space traditionally dominated by IVR systems and human operators. Meta’s 10 million‑user base gives it a sizable launchpad, but adoption will hinge on how smoothly the voice hand‑off works and whether enterprises adopt countermeasures.

Looking ahead, Muse’s paid‑only status may limit early traction, especially as competitors experiment with free or lower‑cost tiers. The upcoming email‑address feature for free users could act as a funnel, encouraging upgrades once users experience the convenience of AI‑driven outreach. Privacy concerns—such as sharing personal account numbers over an AI‑mediated call—could attract regulatory scrutiny, while businesses may deploy AI‑detecting tools to block or flag such calls. Monitoring user churn, token consumption patterns, and any pushback from consumer‑protection agencies will reveal whether voice‑enabled AI assistants become a mainstream productivity tool or remain a premium novelty.