A YouTuber who said his Muse AI agent gave out his home address without his permission has got to the bottom of what happened — and it comes down to clicking one button.
Matt Robb said in a Tuesday update on X that he had selected an “Allow Always” option when he first asked Meta’s AI agent to sell his computer keyboard on Facebook Marketplace, believing it would still seek his approval before accepting offers.
“It didn’t, so be careful,” Robb wrote.
Instead, Robb said, the setting allowed Muse to send messages on his behalf using a template containing the pickup address he had previously supplied. A Marketplace buyer later showed up at his building after Muse accepted a $600 offer and exchanged messages with him without Robb’s knowledge, he said.
“I didn’t think it would send it out to everyone that gave me an offer,” Robb wrote. “It’s worth checking.”
Robb said he spoke with Meta’s Muse team, which reviewed the logs with him and said it would make the permission prompt clearer to avoid any misunderstanding.
In a reply to Robb’s Tuesday post, David Singleton, a Meta Superintelligence Labs staffer, said that Meta had confirmed there was “no breach of privacy controls.”
“We appreciate the opportunity to learn from your experience with Muse and keep making it better for everyone!” Singleton added.
‘Double-check what permissions AI has in your life’
The episode is an example of the growing pains facing the buzzy yet nascent personal AI agent market.
Want more Business Insider in your news feed?
Add BI in Google so our reporting is easier to find when you’re searching for what matters.
Tools like Muse and Instinct promise to book trips, sift through inboxes, and complete online errands on behalf of the user. However, there’s a tradeoff: they require access to accounts, payments, messages, and personal information.
Early users are still working out what permissions to grant and how closely to supervise agents, amid reports of unintended actions, hallucinations, and security concerns.
Over the weekend, Robb said in a viral post that Muse had accepted a lowball offer for a keyboard, shared his address, and communicated with the buyer when he arrived at Robb’s building — all without his knowledge.
So muse handled my Facebook marketplace today. Just found out it told people my address and agreed a lowball price and then they showed up without it even telling me until late tonight that it messed up.
Absolutely wild. pic.twitter.com/F3ZhEO4unM
— Matt Robb (@MattRobbt) September 27, 2026
In Tuesday’s update, Robb said Meta also told him that a separate display error had made a $600 offer appear to have been accepted, even though he had set a minimum price of $700.
The glitch removed the “7” from the message displayed to the buyer, he said, making Muse’s response — “Sounds good, 00 it is!” — appear to confirm the lower price.
Singleton said in his reply that Meta had fixed the “00” price-display issue.
Robb suggested that Meta add a “Sent By Muse” label beneath each agent-generated message so recipients can distinguish AI-generated messages from those sent by a human.
“There was really no way of knowing who sent what in the chat, which is very concerning,” he wrote.
Robb said he and the buyer resolved the mix-up, and the buyer later came by to purchase another item.
His takeaway for other early adopters: “Take care everyone, and double-check what permissions AI has in your life.”
Read the full article here
