Meta on Sept. 8 introduced Muse, a consumer personal AI agent meant to do work — not just chat — across email, browsing, shopping, and longer goals. The U.S. rollout is on iOS, Android, muse.ai, and WhatsApp. Meta says a free tier covers most needs, with subscription plans for people who want to do more.
Muse runs on Muse Secure VM, a dedicated cloud virtual machine with its own browser that Meta says contains the agent and a person’s connected data. A separate Sentinel agent on the same machine must approve outbound internet actions, Meta writes, and Muse asks for permission before sensitive steps like sending email or making a purchase. Meta says Muse has no visibility into passwords or payment methods stored for those flows.
Checkout is wired through Link by Stripe. Meta calls Muse the first AI agent covered by Link’s purchase protections, with one-time-use cards so real card details stay hidden. Stripe says Muse can check out at more than 1 million Link-accepting businesses or use a single-use virtual card elsewhere, and that users approve each purchase total in chat. Shop Pay and 1Password support are listed as coming soon in Meta’s launch post.
Meta also pitches everyday creator-adjacent helpers — for example turning a saved Instagram recipe Reel into a grocery list and remembering friends’ dietary restrictions before sending invites — while framing Muse against the broader agent race with OpenAI and Anthropic workspace tools. Whether people will trust an agent with inbox and wallet access is the open question after launch.