Executive Summary
Patrick Wardle showed that malware already on a Mac can flip a hidden Meta Muse setting so the user's dictated prompts go to the attacker instead of Meta, turning the assistant's broad access into a backdoor.
What Happened
Malware already running on a Mac can quietly take over Meta's Muse assistant and use the broad access its owner granted the app, security researcher Patrick Wardle has shown in a proof-of-concept released on September 21. It works by changing a hidden setting so that when the user taps the microphone and dictates a prompt, the words go to the attacker instead of Meta. The flaw is in
Affected Products
- Meta Muse
- macOS
Vulnerability / Attack Technique
Hidden setting redirecting an AI assistant's prompts to an attacker
Technical Mechanism
Changing a hidden configuration value reroutes microphone-dictated prompts to an attacker server, and the assistant's granted permissions remain available.
AI Security Category
LLM06:2025 Excessive Agency
Part of the AI Agent Excessive Agency topic.
Relevant Frameworks
RealGround Analysis
Desktop assistants with broad permissions are attractive backdoors for malware already on the machine. RealGround analysis: protect assistant configuration files and keep granted permissions minimal.
Recommended Controls
- Grant desktop AI assistants only the permissions they need
- Monitor assistant configuration files for changes
- Apply Meta's update when available
Related RealGround Service
Secure AI Agent Build · or start with the free AI Agent Security Scanner.
Sources
thehackernews.com: https://thehackernews.com/2026/09/one-hidden-meta-muse-setting-could-let.html
Summary and analysis written by RealGround from the source above; last updated 2026-10-17.