Less than two weeks after Meta agreed to a multistate settlement totaling $18 billion in a lawsuit over social media consumer harms, the company introduced Muse, a personal AI agent aimed at helping U.S. consumers with daily tasks and projects. Muse is designed to connect to the apps and services people use in everyday workflows — email, calendars, payments and more — and perform actions on their behalf.
What Muse does and how it works
Unlike chatbots that mainly answer questions or act as conversational partners, Muse is intended to carry out tasks for users. Meta says the agent can send emails, book travel, reduce bills, fill out forms, create plans, convert recipe reels into grocery lists, send party invitations, and make purchases using Link by Stripe for checkout. Shop Pay and 1Password integrations are planned for a later date.
Users decide which apps and services to connect to Muse, adding them one at a time. If a desired service provides a public API, Muse can establish a connection using credentials supplied by the user; if no API exists, Muse can access the service through the browser.
Availability and pricing
Muse will launch in the U.S. on the web at muse.ai, via iOS and Android apps, and through WhatsApp chats, with plans to bring it to Meta’s AI glasses later. The service is free to start but requires a payment card because higher usage will prompt subscription plans. At launch there will be two paid tiers: Power at $20 per month and Maximum at $100 per month. Meta expects most people to remain on the free tier; the app includes a usage meter showing remaining free usage and will warn users when that allotment is exhausted and offer subscription options.
Privacy and security claims
Meta says Muse runs inside a dedicated, secure computer environment it calls Muse Secure VM, which provides privacy and security protections for customer data. A separate Sentinel agent also runs on the same virtual machine but is kept isolated from Muse at the system level. Meta claims that Muse will not have visibility into users’ passwords or payment methods and that user conversations and data will not be shared with Meta’s advertising systems.
The company published technical documentation explaining these protections, but those claims will require independent security analysis.
Trust challenges given Meta’s history
Meta’s past data-privacy incidents and regulatory problems may affect consumer willingness to grant broad access to a personal agent. Notable points in the company’s history include a 2011 FTC finding of deceptive practices around private information, a 2019 FTC settlement that included a then-record $5 billion penalty over eight privacy-related violations, and a 2023 FTC charge alleging violation of the 2019 privacy order. Technical lapses — such as readable-format password exposures discovered in 2019 — and the Cambridge Analytica scandal have also shaped public perception.
Recent legal outcomes contextualize those concerns: alongside the $18 billion multistate settlement announced earlier this month, Meta has faced other judgments and suits, including a New Mexico case that resulted in a $942 million payment and thousands of ongoing personal-injury and school-district cases against social media companies.
User control, customization and next steps
To allay fears, Meta emphasizes opt-in, per-service connections and provides personalization features so users can name and style their Muse agent and adjust how it communicates. Meta also says Muse can continue to operate after a user leaves the app and will learn over time from conversations to offer proactive suggestions.
Muse enters a growing field of so-called agentic AI: other companies and startups are likewise experimenting with browsers, assistant services and chat integrations that perform tasks for users. Whether Muse succeeds will hinge on whether the practical benefits and the company’s technical safeguards are sufficient to overcome lingering doubts about Meta’s data practices. Security experts and regulators will likely scrutinize Meta’s technical claims as consumers decide how much access to grant the new agent.
