Meta Platforms is enhancing security warnings inside its Muse AI agent after bug bounty researchers identified vulnerabilities that could expose user data.
- An ethical hacker discovered a flaw in Meta’s Muse AI agent that could potentially grant unauthorized access to a user’s cloud-based private emails and files.
- Meta is making safety warnings clearer and more prominent when it detects malicious sites.
- The discovery follows a separate Mac-related vulnerability that Meta fixed, highlighting ongoing industry-wide security challenges with autonomous AI tools.
Meta Platforms Inc. (META) is upgrading safety notifications in its Muse artificial intelligence assistant after discovering a security flaw that could let malicious actors access private user files, according to an internal company incident report first reported by The Information.
The vulnerability, uncovered by an outside researcher through Meta’s bug bounty program, threatened a user’s dedicated virtual machine—a cloud-hosted environment containing personal files and emails.
Meta stock eased about 4% on Friday, but gained 13% this week and was on track for its best month in over a decade amid the growing popularity of its Muse AI platform, which topped mobile app charts.
Flaw Downgraded After Initial Review
Internal documents obtained by The Information show Meta initially classified the vulnerability as a “SEV-2” incident, the tech giant’s second-highest severity rating, typically reserved for security issues with significant operational or privacy impacts.
A Meta spokesperson confirmed the flaw was initially miscategorized and subsequently downgraded to “SEV-3”.
To exploit the security gap, an attacker would need to trick a user into asking Muse to summarize or process a link to a compromised webpage. The user would then have to manually select “allow” on a system prompt containing a security notice.
In response, Meta is increasing the visibility of these safety alerts to add a defensive barrier.
Broader AI Security Hurdles
The news comes alongside a separate security finding posted on X, formerly Twitter, by an independent analyst.
That flaw involved local malware on Mac computers that could theoretically redirect Muse’s audio recordings to external servers and compromise user account access. Meta officials stated that the Mac-related issue posed a low operational risk and confirmed that a fix has already been deployed.
Meta Stock: Retail View
Retail sentiment on Stocktwits was ‘bullish’ with ‘high’ message volumes.
META stock has gained about 14% year-to-date.
For updates and corrections, email newsroom[at]stocktwits[dot]com.<