Metas Muse AI leaves users vulnerable to identity theft: report | Latest Tech News
It’s an AI agent of chaos.
Meta’s allegedly ironclad AI assistant Muse sparked privateness issues after it was caught exposing users’ delicate and business info.
Ahead of the Muse’s release last month, Meta CEO Mark Zuckerberg had burdened the “privacy and security” of the cyber secretary, which executes duties on the users’ behalf, from sending emails to reserving journey, Futurism reported.
However, an investigation by security firm Claroc discovered that the supposedly watertight tech was something but. The company’s investigators reportedly found a glitch that allowed them to access pages from over 6,800 hyperlinks Muse had created so users may share pages with each other, Ars Technica reported.
An investigation discovered glitches in Meta’s AI assistant Muse that allowed them to access delicate personal and business data. @alexandr_wang/X
Mark Zuckerberg, CEO of the Facebook-owned Meta Group, unveils a small gadget called Muse Charm designed to present access to the company’s proprietary AI software program, Muse, at the Meta Connect developer convention in Menlo Park, United States, Wednesday, Sept. 23, 2026. ZUMAPRESS.com
These included loan purposes with the requested quantity, bank transaction particulars, a real estate investor’s inside efficiency report, and a company income doc meant for the CEO.
There was even a company’s full-year trip calendar itemizing the PTO dates of dozens of workers.
Worst of all, these secrets and techniques had been unveiled with the push of a button. “If you had asked us beforehand, we would have guessed that gathering these would take real effort,” the researchers wrote. “It took one query.”
To access this intimate data, the investigators reportedly exploited Muse’s share operate, which permits users to share searches, conversations, pictures, and other content among one another.
A meme that includes Jolly’s, Muse’s mascot. Alexandr Wang/X
This content is accessed via AI-generated web hyperlinks that sometimes incorporates a long, cryptic string of characters that theoretically make it laborious to hack.
However, the Claroc crew discovered Muse’s URLs apparent as they all used the area identify cf.metaaiusercontent.com as properly as the AI-generated dialog title alongside the textual content sequence.
“If you had asked us beforehand, we would have guessed that gathering these would take real effort,” the researchers wrote. “It took one query.”
Examples included home-sale-budget-9f21c8a3.cf.metaaiusercontent.com and loan-application-draft-a4c9e2.cf.metaaiusercontent.com.
“We call these GhostShares: pages that open for people who were never handed the link, and that those people can find on their own,” wrote Claroc.
In mild of the oversight, Meta has shut down the characteristic. “We opened the Muse bug bounty program so researchers could raise different scenarios and help us continue improving the product,” company reps told the Post. People have the power to share particular Muse artifacts, which creates a web page that will be accessed by those with the hyperlink. However, we’ve paused this performance while we review this report and contemplate adjustments to how these hyperlinks will be found.”
However, this wouldn’t be the first time somebody has allegedly found a privacy-violating glitch in the Meta-verse.
Last week, Meta glasses wearers claimed that the cutting-edge tech specs had been listening to them even when set to sleep mode.
“Meta AI does not respond to me, and I never intentionally activated it,” lamented one Redditor. “I’ve observed the white notification mild activating while I’m talking usually.
Stay informed with the latest in tech! Our web site is your trusted source for breakthroughs in artificial intelligence, gadget launches, software program updates, cybersecurity, and digital innovation.
For contemporary insights, knowledgeable coverage, and trending tech updates, go to us recurrently by clicking right here.
