Meta's Muse AI agents are gaining traction in the US, with the app passing 5 million downloads last week. That makes it worth reviewing what information Meta gathers to power Muse's recommendations. Social Media Today looked at the issue based on reporting from Wired.
According to Wired, security researchers extracted Muse operating-system elements, including instructions on how the app should gather and use data. One instruction appears to let Muse create a page for every person in the user's life. The process runs hourly and compiles data on family, partners, friends, colleagues, collaborators and people the user follows, stored in Muse's "memory," which is structured text files.
The instructions say Muse could track where people live, what they do, recurring threads such as an apartment move or a shared savings goal, and "dates that matter" like birthdays and anniversaries. A person's history could include the trip in March, an argument that got resolved or a milestone last week. For a tool meant to coach relationships, that makes sense, but it is also a lot of personal information logged for each user.
Meta makes 97% of its revenue from advertising, so Social Media Today worries that some of these insights could feed ad targeting, which could exploit psychological vulnerabilities, shape political content or keep users hooked. A leak is another risk. Muse downloads are growing faster than ChatGPT did at the same stage, and Meta has often countered privacy concerns with utility.
As personal agents spread, advertisers need to design for data ethics and consumer trust at the same time. If agent-derived signals ever reach ad products, first check whether you can explain which data any targeting is based on. TikTok is expanding its own agentic ad tools in the same week, so the platform AI race will continue. For background on Muse's launch, see Meta Launches Muse — Its Latest Bet That People Want a Digital Assistant. For help with data-driven marketing, see our services.
According to the instructions Wired reported, it can log where family, friends and colleagues live, what they do, recurring life threads, important dates like birthdays and recent events.
Meta earns 97% of its revenue from advertising, so some of these insights could be used for ad targeting.
The app passed 5 million downloads last week and is reportedly growing faster than ChatGPT did at the same stage.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
TikTok’s Road to Black Friday series shows how top TikTok Shop sellers prep for BFCM: Princess Polly’s 6,000 creator videos a month, Dossier’s sampling engine and SharkNinja’s LIVE strategy.
Meta shared how small businesses use its Muse AI agents: analyzing ad performance, building campaigns through conversation and automating workflows with Shopify, Canva and HighLevel. Prompts included.
Reddit outlined its U.S. midterm election plans, including the Up the Vote hub, AI detection that cut spam exposure 20% and removes nearly 2 million fake upvotes daily, and AMA rules for political ads.