Meta’s Muse AI Agent Faces Privacy and Security Backlash Amid Rapid Growth
Meta’s Muse AI agent has surged to the top of app stores, but faces severe criticism over privacy and security. Reports reveal it builds detailed dossiers on users and their contacts, while internal documents show rushed fixes for critical virtual machine vulnerabilities before launch.
Key points
- Muse has over 5 million downloads and is the top free app on iOS, driving a 20% stock increase for Meta.
- Internal instructions show Muse updates hourly dossiers on users and their social networks, tracking relationships and behaviors.
- Meta rushed to fix 'KVM escape' vulnerabilities before launch, with engineers fearing a massive data breach.
- Meta offers free access to build habits, with subscription plans at $20 and $100 per month.
- Security researchers found zero-day exploits allowing unauthorized access to user data and Instagram followers.
Background
Meta launched Muse at its Connect 2026 event, positioning it as a core AI strategy to diversify away from ad revenue. The company also introduced the Muse Charm hardware device and partnered with retailers like Walmart, though Amazon and Insurify blocked access due to security concerns.
How outlets are covering it
Futurism and Time Magazine emphasize privacy risks, noting Muse tracks non-users and shares anonymized insights across virtual machines. 404 Media highlights security flaws, including a 'mad dash' to fix KVM escapes that could expose Meta’s internal databases. CNBC focuses on commercial viability, questioning if users will trust Muse with personal data after a $17 billion settlement over child mental health harms. Meta defends its design, claiming users control their data and that virtual machines are isolated, though critics argue the architecture is inherently risky.
Why it matters
Muse’s rapid adoption raises critical questions about AI agent security and privacy. If Meta cannot ensure robust isolation, a breach could expose millions of users’ sensitive data, undermining trust in AI assistants and potentially triggering regulatory scrutiny.
What to watch
Meta plans to introduce encryption for Muse data later in 2026. The company will continue monitoring for vulnerabilities, while competitors like OpenAI and Google face similar challenges in securing agentic AI systems. Consumer adoption may slow if security concerns persist.
- Meta’s Muse Is Spying on All Your Friends and Family Futurism
- Meta's Muse assistant tops app charts. Now it needs to become a habit CNBC
- Meta’s Muse AI Agent Is Building a Dossier On You Time Magazine
- Citigroup bets on Meta's Muse as AI's 'front door to the internet' Reuters
- Meta Rushed to Fix Muse ‘VM Escape' Vulnerability Soon Before Launch 404 Media
Want the full story? Read the original reporting
Read on Futurism