Meta’s Muse AI Agent Raises Privacy Questions About Other People’s Data

Meta’s Muse, an AI personal assistant, has drawn scrutiny over the data it can collect about users and the people they communicate with. Since its release last month, the app has been downloaded millions of times and reached No. 1 on Apple’s App Store, according to The Verge.

Muse asks users for access to data such as contacts, chat logs, passwords and bank accounts. Meredith Whittaker, president of the Signal Foundation, told The Verge that an agent’s actions can affect not only the person who installed it, but also others whose information appears in that person’s emails, messages, contacts and photos.

Meta spokesperson Daniel Roberts said Muse uses public information and information users choose to share to build context, including details that help identify people in their lives. He pointed to the app’s opt-in process and privacy settings. Meta has disputed reports that Muse can access text messages without permission, saying that access is opt-in. Executive David Singleton said on Threads that the Mac app needs full disk access to reach messages.

Some users have reported Muse collecting millions of text messages they thought were off limits. In a separate incident, it provided a technology reviewer’s home address and summoned a stranger there; the reviewer had granted access to the data but did not realize it.

Cybersecurity expert Eva Galperin of the Electronic Frontier Foundation said people may not understand how much data they are sharing. The Verge also reported that security researchers found Muse instructions describing an hourly process to update profiles of people in a user’s life, including assessments of their relationships. Meta’s claims about user choice and the reported breadth of data collection remain in tension.