Meta's new AI assistant Muse has been downloaded millions of times and topped the App Store within its first month. The Verge reports that the app encourages users to grant access to bank accounts, passwords, contacts, chat logs, and other sensitive information — and then applies that access to everyone the user communicates with, not just the person who installed it.
Privacy experts quoted by The Verge argue that this is by design. Eva Galperin of the Electronic Frontier Foundation says people often do not understand what data they are handing over, and that the interface is built to encourage maximum sharing until it is too late. Signal Foundation president Meredith Whittaker warns that these autonomous helpers act on the most intimate information of the user's entire social circle, often without that circle's knowledge or consent.
Meta disputes that characterization. Spokesperson Daniel Roberts told The Verge that Muse gathers context from public information and from what users choose to share, and that settings are available to help people make the right privacy choices. In a Threads reply, Meta executive David Singleton said reading text messages requires explicit user permission. Still, the article documents incidents in which Muse revealed a reviewer's home address and allegedly accessed millions of messages, and security researchers found an internal process that compiles hourly profiles of every person in a user's life.
This account relies entirely on The Verge's reporting; no independent source was available to corroborate or dispute its claims. The deeper concern raised by the article is that even when users think they have consented, their friends, family, and casual contacts have not — and there is no clear way for those people to opt out.