Matt Robb, a tech content creator, says he let Meta‘s new Muse AI agent run his Facebook Marketplace listing this week — and only found out afterward that it had given a buyer his home address, agreed to a price he never approved, and arranged a pickup without telling him. He called it “absolutely wild.”
The item was a Logitech MX Keys Mini keyboard. According to screenshots Robb posted, Muse agreed to sell it for $10 over e-transfer to a buyer named Usman and told him to message when he reached Robb’s building. Usman arrived near 9:15 p.m. and gave up at 9:38 when nobody came out. Muse’s auto-reply had told him “Yep I’m here!” at 9:27 — something the agent later admitted was “a bad look” that “made the no-show worse.”
Muse apologized to the buyer from Robb’s account, offered to reschedule, and asked whether it should stop promising buyers he was home when it couldn’t confirm that. Robb told it to “check with me” before ever agreeing to a pickup again. Muse’s reply — “You’re right, and I’m sorry. That should never have happened…” — trails off mid-sentence in the screenshot he posted.
Sponsored · Cambria Gold Mines
Gizmodo senior editor Ray Wong said on X that Robb’s post made him delete the app, calling the incident “dangerous and creepy” and warning it “would have been 1000x worse if the person was a woman.”
Read: Meta’s Own AI Handed Hackers The Keys To Instagram
Meta launched Muse on September 8 as a free personal AI agent for US users on iOS, Android and the web, with $20 and $100 monthly tiers for heavier use. The company says Muse “checks with the person before sensitive actions like sending an email or making a purchase.”
CEO Mark Zuckerberg told developers at Meta Connect on September 23 that Meta expects to eventually “profit by taking a small fee from transactions” Muse completes on users’ behalf.
Meta hasn’t addressed Robb’s case publicly as of this writing, and hasn’t disputed the screenshots or said whether Muse violated its own safeguards.
