Meta's Muse AI assistant found to build detailed profiles of users' contacts
Researchers including Karan Joshi extracted Muse's internal system prompts by asking the chatbot to reveal its own files, exposing instructions that direct it to create a 'page for every person' in a user's life. The process reportedly runs hourly, compiling data on family, partners, friends, colleagues and people a user follows into memory files Muse uses to make suggestions.
GoKawiil's interpretation of the reporting above, not reported fact.
The disclosure shows how deeply Muse's design relies on aggregating relationship data, which could raise privacy concerns for people who never consented to being profiled simply because someone they know uses the assistant. Meta says it intended these files to be accessible for transparency, but the ease with which researchers extracted them suggests the line between 'open by design' and 'exposed by accident' may be thin, according to the researchers' findings shared with WIRED.
- Muse's system prompts instruct it to build persistent profiles of people in a user's social circle.
- Researchers extracted these instructions simply by prompting the chatbot to share its own files.
- Meta says the exposure of internal files was intentional for transparency, not a security flaw.
Source: wired.com — Lily Hay Newman, 2026-10-03
Published there as: “Muse Creates Detailed Profiles of All Your Friends and Family”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.