Meta’s Muse AI agent collects data on users—and non-users—raising privacy concerns

Meta’s AI agent Muse, launched in September 2026, has become a top download on Apple’s U.S. App Store—but its data collection practices are drawing scrutiny. Researchers say the app updates detailed files on users and non-users alike, raising privacy and security concerns.

Muse’s Data Collection Practices

Meta’s AI agent Muse, launched on September 8, 2026, has quickly climbed to the top of Apple’s U.S. App Store chart for free iPhone apps. However, its rapid adoption has been accompanied by concerns over how it collects and updates data about users and the people in their lives.

According to reporting by Gizmodo, Muse creates and maintains detailed files on individuals connected to its users, even if those individuals do not use the app themselves. These files include names, relationship details, recurring topics of conversation, and important dates. The agent is designed to update this information hourly, based on system instructions uncovered by independent AI safety and security researcher Karan Joshi.

Security and Platform Restrictions

Muse’s data-handling practices have already led to friction with major platforms. Amazon blocked the AI agent from shopping on its site in September 2026, stating that Meta had not notified the company in advance about its integration plans. While the specifics of Amazon’s concerns were not disclosed, the move highlights broader tensions over how AI agents interact with third-party services.

Meta has also faced security challenges with Muse. Gizmodo reported that the company patched a zero-day vulnerability in the Mac version of the agent that could have allowed malware to hijack its functionality. The flaw was addressed, but its discovery underscores the risks of deploying AI agents at scale.

How Muse Stores and Protects Data

Meta claims that Muse stores user data in dedicated secure virtual machines (VMs), which are isolated and inaccessible to other users’ agents. The company states that the agent gathers context from public information and data that users choose to share, but the extent of this collection has raised questions about consent and privacy.

The agent’s system instructions, as revealed by Joshi, indicate that Muse can maintain a history of relationships, including events, milestones, and the perceived closeness between users and their contacts. This level of detail has sparked concerns about the potential for unintended exposure or misuse of personal information.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

Muse’s rapid rise to the top of the App Store reflects the growing appetite for AI agents that manage personal and social data. However, the revelations about how Muse collects and stores information—including about people who don’t use the app—highlight a critical tension in AI development. Founders and operators should note that users and regulators are increasingly scrutinizing how AI tools handle relationships, context, and consent.

Meta’s decision to store data in secure virtual machines may address some security concerns, but the broader question of ethical AI design remains unanswered. For startups building similar tools, this story underscores the importance of transparency, user control, and proactive collaboration with platforms like Amazon to avoid friction. The patching of a zero-day vulnerability also serves as a reminder that even well-resourced companies must prioritize security in AI agent development.

Key takeaways

• Meta’s Muse AI agent has gained widespread adoption but is under scrutiny for privacy practices.
• The app creates and updates detailed profiles of users and non-users based on shared data.
• Amazon blocked Muse from shopping on its site due to concerns about data handling and notifications.
• Researcher Karan Joshi revealed system instructions showing hourly updates to relationship data.
• Meta addressed a zero-day vulnerability in Muse’s Mac version, highlighting ongoing security risks.

FAQ

What is Muse, and why is it controversial?

Muse is an AI agent developed by Meta, launched on September 8, 2026. It has sparked controversy due to its practice of collecting and updating detailed information about users and the people in their lives, even if those individuals do not use the app. Concerns center on privacy, consent, and the security of stored data.

How does Muse collect and store data?

According to researcher Karan Joshi, Muse creates files on individuals that include names, relationship details, recurring topics, and important dates. The agent updates this information hourly and stores it in dedicated secure virtual machines, which Meta claims are isolated and inaccessible to other users.

Why did Amazon block Muse from shopping on its site?

Amazon blocked Muse in September 2026, citing concerns over Meta’s failure to notify the company in advance about the AI agent’s integration with Amazon’s shopping platform. While the exact issues were not disclosed, the move reflects broader tensions over data handling and platform policies.

What security issues has Muse faced?

Meta patched a zero-day vulnerability in the Mac version of Muse that could have allowed malware to hijack the agent. The discovery of this flaw highlights the security risks associated with deploying AI agents at scale.

How can users control what Muse collects about them?

Meta has not publicly detailed options for users to limit Muse’s data collection about non-users. However, the company states that the agent gathers context based on public information and data that users choose to share. Concerns remain about the lack of transparency and consent for individuals who do not use the app.

Related on Lazyfounder

• Amazon
• Apple
• More ai news
• OpenAI Fires Three Employees Amid AI Safety and Data Mishandling Concerns