Why Meta Built Muse, And Why Some People Don’t Trust It Yet
Meta's new AI agent can shop, plan and message for you. But its short life has already shown both its usefulness and its risks.
Meta Built Something That Can Spend Your Money For You
An app that quietly books your flight, cancels a subscription you forgot about, and buys your movie tickets while you do something else. That is what Meta is selling with Muse, its AI agent that launched September 8 and overtook ChatGPT as the top free app on Apple’s US store within two weeks.
Why Meta Built It
Muse runs on Meta’s own Muse Spark model and connects to your email, calendar, and shopping apps. Zuckerberg calls it his path to “personal superintelligence,” and he is not waiting around — Connect this week brought a keychain-sized device called Muse Charm, an avatar you can talk to, and shopping deals with Walmart, GameStop, Gap and Wayfair. Notably, Amazon has blocked Muse from its store, a sign of how seriously rivals take Meta’s push.
The Side Effects
The upside is real: Meta’s stock has climbed nearly 25% since Muse launched, and users say it genuinely saves time. But Muse’s short history already shows the risks of handing an AI your money and personal data. An earlier feature, Muse Image, was pulled within days after it used people’s public Instagram photos to generate new images without real consent, drawing criticism from SAG-AFTRA and talent agency CAA. Reuters also reported internal Meta tests found security flaws and “failure modes” that made the assistant unreliable.
None of this has caused confirmed harm yet. But Meta is asking people to trust it with bookings, payments and health data — the same company that recently paid $17 million over failing to warn the public about harm on its platforms. Whether that trust holds may decide if Muse becomes essential or just another app people quietly stop using.
Get the latest tech news on WhatsApp
Comments are closed.