Meta’s new personal assistant, Muse, is building detailed profiles of every person in your life, including family, friends, and colleagues.
Millions of users have downloaded the AI agent and connected it to bank accounts, messages, and health data. As consumers adopt the tool, researchers have extracted internal files to reveal how the system organizes information.
Meta states it intended for these files to be accessible to ensure transparency. The documents show how the system responds to prompts regarding sensitive or politicized topics.
Independent researcher Karan Joshi extracted an extensive array of instructions by asking Muse to copy and share its own software files through the regular chat interface. He subsequently shared his findings with WIRED.
One instruction appears to allow Muse to create a page for every person in the user’s life. This hourly process compiles data on family, partners, friends, colleagues, collaborators, and people you follow.
The system uses structured text files to collect information about relationships. It can then make suggestions, such as advising on how to improve a connection or suggesting a coffee shop for a friend.
While AI chatbots have tracked social information for years, Muse is notable for its setup given Meta’s history and access to social network data.
“What it seemed like to me—from all these prompts, system skills data, and things that they’re feeding into Muse—is that they want to understand your relationships that you have with real people,” Joshi says. “They’re trying to know you like a friend, which is honestly pretty creepy.”
Muse’s documentation states a page may start sparse and fill out over time. Sections could include Facts, History, The relationship, In common, Open threads, and Strengthening.
Instructions say the model should only use available evidence. Invented details are considered worse than an empty page.
The documents note where people live, what they do, and recurring threads such as an apartment move or a shared savings goal. The system could also record dates that matter, like birthdays or anniversaries.
A person’s history might include backstory such as a trip in March, an argument that got resolved, or a milestone from last week.
Instructions focus on recording relationship details such as how close people are, what the relationship is built on, how they act with each other, and what it seems to need right now.
The Strengthening section suggests ways to improve a relationship. Examples include a reason to call, a date worth remembering, something they said to circle back on, or a way to be there for them that matters.
“We are giving AI systems much more information about us than we are getting information from them,” says Carissa Véliz, an associate professor at Oxford’s Institute for Ethics in AI. “It’s not only what we explicitly tell them, but what they can infer from us—correctly or incorrectly, both concerning for different reasons—and what they can piece together from other sources of data.”
Muse is architected so each individual user has their own dedicated virtual machine. This stores user data and context. The machine is inaccessible to other agents. Users can wipe memories or disconnect external services at any time.
Meta says Muse seeks human confirmation before completing actions like sending an email or making a purchase. An audit log allows users to see all of the agent’s activity and future plans.
“For any agent to be useful and actually help you achieve your goals, it needs to have context about you and those you interact with,” Meta spokesperson Daniel Roberts told WIRED. “Muse gathers that based on public information and from what you’ve chosen to share, which is how it remembers the person who sent you an invoice is in fact the plumber who you previously hired to complete work in your bathroom or which flowers your spouse said they liked best.”
It is common for AI assistants to incorporate memory features designed to allow greater personalization. Miranda Bogen, the director of the Center for Democracy and Technology’s AI Governance Lab, says Muse appears to have more emphasis on relationships and personal contacts than rival systems.
These tools tend to offer transparency and editing capabilities. Muse does as well. Bogen notes that agents encourage people to proactively share more data with them versus focusing on culling data.
“These [AI assistant] tools are actively soliciting users to plug their whole lives in—their emails, calendars, financial institutions, everything in order to be helpful assistance,” Bogen says. “That’s dramatically more information than people might have otherwise given to some of these companies. The breadth of access to information that these tools have will lead to a ballooning of what they know about users.”
What it means
The shift is from tools that answer questions to systems that map your social circle. Users are asked to connect their entire digital footprint to the assistant. This grants the software access to a breadth of information that exceeds what people typically share with companies.
