Meta announced Muse AI on September 8, 2026, a personal agent app called Muse that the company says is built to automate routine digital tasks and interact with a user’s other apps.
In its official blog post, Meta described Muse as “a secure, private personal AI agent” that people can message in a standalone Muse app or in WhatsApp to ask it to run errands, manage schedules and carry out multi-step tasks, according to Meta’s announcement.
Reuters reported that Muse is designed to connect to a person’s email, calendar, payments, health, shopping and smart-home apps so it can perform actions such as sending emails and making payments, and that Meta said users can review and approve actions before they happen.
News coverage and Meta’s product page list examples of what Muse can do: send emails, book travel reservations, fill out forms and make purchases across connected services, with the company emphasizing that users choose which apps and permissions to grant.
CNBC reported Muse will be available in free and paid tiers, with subscription options described in early coverage; the outlet said Meta outlined $20 and $100 monthly tiers depending on usage levels.
Tech writers noted the move places Meta into a crowded market of personal AI agents and raised questions about trust and safety. Wired and TechCrunch highlighted Meta’s privacy and safety framing, saying Meta built Muse to operate in a secure cloud and to give people controls over access.
Policy context is already shaping reactions: Axios and other outlets pointed out the announcement came amid ongoing policy debates, and several outlets cited U.S. lawmakers’ recent push for AI safety rules that could affect services like Muse.
One practical early friction was reported in other tech coverage: Amazon moved to block third-party AI tools from making purchases on its site, a step that could limit some Muse capabilities for shopping, according to technology reporting on industry reactions.
Meta’s Muse launch follows a broader industry trend of companies building agents that act inside multiple apps; coverage named Muse as Meta’s latest major product push in AI and quoted company material explaining how users must grant permissions and can review tasks before execution.
Sources
- Meta (about.fb.com) — announcement describing Muse as a private personal AI agent and where it can be used.
- Reuters — reporting that Muse can connect to email, calendar, payments and other apps and perform actions such as sending emails and making payments.
- CNBC — details on Muse’s subscription tiers and pricing mentioned in early coverage.
- Wired — context on privacy framing and how Meta positions Muse’s safety controls.
- TechCrunch — analysis of consumer trust questions and product capabilities.
- Axios — commentary linking the launch to U.S. AI safety debates and legislation context.
- Eciks.org — technology coverage reporting Amazon blocked some Muse shopping actions (internal link used earlier for context).
