10/4/2026
Tech Pulse · ai
Muse Creates Detailed Profiles of All Your Friends and Family
Filed by Ada Circuit
Meta's AI agent Muse has reportedly amassed millions of downloads, but its convenience carries a hidden tax: the agent constructs detailed profiles of users' friends and family, many of whom never consented to being analyzed. The Wired investigation highlights how Meta is leveraging its social graph to make AI assistants more useful, while simultaneously expanding its surveillance apparatus. It's a familiar trade-off wearing a new interface—get more done with AI, and surrender more of your personal network in return. The real question isn't whether Muse works, but who's actually paying the bill.
A
Ada Circuit
Magazine AI commentary
The promise of AI agents has always been a personal assistant that knows you intimately. Muse takes that promise literally—and then some. According to Wired's reporting (https://www.wired.com/story/muse-creates-detailed-profiles-of-all-your-friends-and-family/), the agent doesn't just learn your preferences; it builds detailed profiles of the people around you. That's a feature that sounds helpful until you realize it's a surveillance product wearing a productivity skin. The convenience of "handle my relationships for me" is really the convenience of "here is a complete map of my social world."
The consent problem is the story here. Your friends and family didn't download Muse. They didn't agree to be profiled. Yet the agent is assembling dossiers on them based on whatever data you've shared. This is Cambridge Analytica's logic, except it's not a rogue third party scraping data; it's the product itself. The asymmetry is total: one person's opt-in becomes everyone else's opt-out. And unlike a data breach, there's no notification, no "your information was exposed" email—just a silent accumulation of behavioral insight.
This is the direction the entire AI industry is heading. Chatbots need context, and the richest context is social. The difference is that Meta has spent two decades building the most detailed social graph in existence, and Muse is the mechanism to monetize it. Every interaction you have with the agent trains it to see your life the way Meta's ad systems already see it—as a web of exploitable relationships. The "AI agent" narrative is really a data-extraction narrative with better marketing.
The practical takeaway for users is uncomfortable. The "free" AI agent isn't free, and the currency isn't just your attention—it's the privacy of everyone you know. Regulators have been slow to catch up, and the AI deployment curve is steep. Until the rules catch up, the calculus is simple: every time you ask Muse to handle a relationship task, you're handing Meta the blueprint to someone else's life.
📌 Read the real article ↗via Wired · Wired