Less than two weeks after Meta agreed to an $18 billion multistate settlement over consumer harms linked to social media, the company on Tuesday launched Muse, a personal AI agent designed to handle everyday tasks like sending emails, booking travel, and making purchases. The rollout, which begins in the U.S., represents Meta’s most ambitious consumer AI bet yet — and one that hinges on a commodity the company has often struggled to retain: user trust.
Muse connects to a user’s apps and services — email, calendars, payments, health, fitness, smart home, dining, shopping, music, and events — to act on the user’s behalf. It’s available on the web at muse.ai, through iOS and Android apps, and in WhatsApp chats. Meta says it will soon come to its AI glasses.
Also read: US military disables ad tracking on troops' devices to counter location-based threats
The agent is powered by Meta’s Muse Spark model and includes built-in connectors for several services. If a service lacks a dedicated connector but has a public API, Muse can use credentials the user provides. When no API exists, Muse can operate through a browser.
How Muse works — and what it asks for in return
Muse is free to start, but requires a payment card to begin. Meta says most users will stay on the free tier, with paid plans — Power at $20 per month and Maximum at $100 per month — offering more usage. The app includes a usage meter that shows how much of the free allowance remains, and it warns users before charging.
Also read: Judge blocks X rival from using Twitter name, but allows ‘Tweet’ for now
Users can decide which apps to connect, doing so one at a time, and can customize Muse by naming it, choosing an avatar, and setting communication preferences. The agent can continue working after the user leaves the app and may make unprompted suggestions as it learns from conversations.
Meta says purchases made through Muse are processed via Link by Stripe, which offers purchase protections — a feature designed to ease concerns about letting an AI check out on a user’s behalf. Shopify’s Shop Pay and 1Password integrations are expected soon.
Muse is not alone in this space. Other companies are exploring similar agentic AI, including Google’s Gemini Spark and Anthropic’s Claude Cowork, which can perform tasks in a browser or across apps. Apple has been integrating AI into iMessage and other chat platforms.
Privacy promises vs. a troubled track record
Meta claims Muse operates in a dedicated, secure virtual machine with its own browser, called Muse Secure VM. A separate Sentinel agent runs on the same VM but is isolated at the system level, according to the company. Meta says Muse cannot see passwords or payment methods, and that it does not share conversations or data with its ads systems.
These claims are detailed in a technical post released alongside the product, but security experts will need to scrutinize them. The company’s history suggests why such scrutiny matters.
In 2011, Meta settled with the FTC over charges it deceived consumers by making private information public without approval. In 2019, the FTC fined Facebook $5 billion for eight privacy violations. In 2023, the FTC charged Meta with violating that same privacy order. Meta also discovered in 2019 that millions of users’ passwords had been stored in readable format, and the Cambridge Analytica scandal — where data from millions of users was collected without consent — still shapes public perception.
More recently, Meta has faced repeated congressional testimony over its handling of minors’ data and safety. The company settled with 29 states in August, paid $942 million in a New Mexico lawsuit over child harm, and still faces thousands of personal-injury and school-district cases.
Can a personal touch overcome a trust deficit?
Meta is betting that personalization will help. Users can give Muse a name, pick an avatar, and adjust how it communicates — a design choice that may make the agent feel less like a faceless corporate tool and more like a helpful companion.
But the fundamental question remains: will consumers hand over access to their email, calendar, and payment apps to a company with Meta’s track record? The utility of a true AI assistant is clear, but so are the risks. If Meta’s history has taught users anything, it’s that promises of privacy are not guarantees.
Muse’s success will depend not only on how well it performs tasks, but on whether Meta can demonstrate — through transparent practices and verifiable security — that it has changed. The company says it has built safeguards into Muse’s architecture, but trust is earned over time, not announced in a product launch.
Disclaimer: This article is for informational purposes only and does not constitute financial advice. The technology and AI markets are volatile and uncertain; readers should conduct their own research before making any decisions.
Warning: Attempt to read property "term_id" on false in /www/wwwroot/stockpil.com/wp-content/themes/flex-mag/functions.php on line 998
