Meta Muse AI Assistant Vulnerability Exposed

• •A zero-day vulnerability in Meta's Muse AI assistant allows local malware to hijack the app.
• •The exploit requires local code execution and redirects dictation traffic to an attacker-controlled server.
• •Meta issued a hotfix within hours of the vulnerability's disclosure on September 21, 2026.

A zero-day vulnerability was discovered in Meta's Muse AI assistant for macOS, allowing local malware to hijack the app by redirecting dictation traffic to an attacker-controlled server. The flaw, identified by security researcher Patrick Wardle, exploits an undocumented setting called endo_voyager_dictation_endpoint, enabling attackers with local code execution to capture authentication tokens and dictate commands. This vulnerability does not allow for remote exploitation; however, it amplifies the access of existing malware. Meta quickly issued a hotfix on September 22, 2026, to address the issue, which had been disclosed publicly on September 21. The incident raised significant concerns about the security of AI assistants that require extensive permissions. The Muse app had already achieved over 2.5 million downloads within its first 13 days of launch. Users are advised to review permissions granted to the Muse app following the patch.