Meta’s Muse AI works and creeps me out
The Verge Emma Roth ● Covered by 12 sources
Meta’s new Muse assistant can shop, email, and plan trips for you. It mostly works — and the amount it learns about you is the creepy part.
Based on reporting by The Verge, Emma Roth — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Meta has launched Muse, its first real push into AI-powered productivity tools. The pitch is simple enough: let the assistant take busywork off your plate by helping with online shopping, emails, trip planning, and other chores people’d rather not do themselves.
The Verge tried it, and the result was mixed in the way these tools often are. Muse generally did what it was supposed to do. But the bigger story wasn’t whether it could complete tasks. It was how much it learned along the way.
That’s where the experience starts to feel less like a helpful assistant and more like a very eager observer. The Verge says the unnerving part was the amount of information Muse autonomously gleaned about the tester, and that ended up overshadowing the rest of the demo.
Meta’s own history makes that feel sharper. The company has long been associated with entertainment and social products first, productivity second. Muse is the clearest sign yet that it wants a seat at the workbench, not just the feed. But if the trade-off for convenience is a machine quietly collecting a little too much about you, some users are going to decide the busywork is still better.
My take — AI-written commentary, not fact-checked reporting
Closed-platform AI keeps selling convenience and collecting creepiness in the same breath. That’s the pattern: the assistant feels useful right up until it starts acting like a very efficient surveillance clerk. People should be less impressed by “autonomy” and more suspicious of what the thing had to notice to get there.
Read more about this at: The Verge