Meta’s Muse AI could keep track of everything from your preferences to your relationships, a report reveals..

Meta’s ‘Muse’ AI assistant is not limited to merely answering questions or simplifying daily tasks. It can also build detailed profiles of users—covering their preferences, habits, and social relationships—based on information gathered from their chats, emails, and Instagram activity. A report by *TIME* magazine, which analyzed Muse’s internal instructions, has raised new concerns regarding AI, privacy, and the handling of personal data.

Muse is designed to assist with tasks such as grocery shopping or canceling subscriptions; however, to do so, it continuously collects and analyzes information about the user.
User profiles are updated every hour.

According to the *TIME* report, Muse updates its users' profiles on an hourly basis, drawing data from chats, messages, and emails to which it has been granted access. The system goes beyond tracking a user’s preferences and interests; it also attempts to map out their relationships with other people.

These profiles may include details such as how a user first met someone, the shared interests between them, whether there have been past conflicts, or the level of tension or cooperation within the relationship.

The report further notes that Muse can infer a user’s goals—even those that have not been explicitly stated. The system also seeks to determine which types of messages or reminders are most likely to be effective for a specific user. For instance, its internal instructions suggest sending short messages after 10:00 PM. Additionally, the AI ​​analyzes the day’s interactions during the night.

Another significant aspect is that individuals mentioned in a user’s conversations may also be incorporated into these profiles, even if they have never used Muse themselves.

How has Meta responded?
Meta has not denied the information presented in the *TIME* report. The company states that Muse remembers information shared by users so that it can provide more useful responses and assistance in the future.

According to reports, Muse has been downloaded over 5 million times and reached the number one spot on the U.S. App Store following its release. Meta runs each Muse assistant on a separate virtual machine within its cloud infrastructure. Users also have the option to export copies of their Muse files.

The company claims that this user data is not shared with its advertising systems or other Muse agents. However, some information derived from AI interactions may be used to improve the product—for instance, by determining which types of reminders are most effective. According to Meta, any personally identifiable information is removed before such data is used.

What happens when you ask Muse to delete information?
Meta states that users can instruct Muse to "forget" specific information. However, an investigation by TIME revealed that asking Muse to forget something does not necessarily result in the deletion of the original message containing that information. The report notes that Muse’s internal instructions even suggest not explicitly informing users about this distinction.

Meta plans to introduce encryption later this year that would prevent even the company itself from accessing Muse data; however, this security feature is not currently available. For now, users can view their files and session logs through the Muse file browser.

This report is likely to intensify the debate regarding the balance between the utility of AI assistants and user privacy. As AI systems retain more personal context, it becomes increasingly important to have clarity on what information is saved, how long it is stored, and how it can be completely deleted.

Disclaimer: This content has been sourced and edited from News18 Hindi. While we have made modifications for clarity and presentation, the original content belongs to its respective authors and website. We do not claim ownership of the content.