🔍 Read the full analysis: What Non-Muse Users Should Know About Meta’s AI Agent on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Meta has expanded its Muse personal AI agent to Singapore, while reporting has raised questions about how it stores information about people who never signed up. Meta says users control Muse’s access; its policy also says Muse may store information about other people. Non-users have limited controls over what someone else’s agent can access.
Meta announced the expansion of its Muse personal AI agent to Singapore on October 27, as reporting has raised questions about its ability to store information on people who never signed up. Meta says each user controls what their agent can access, but its privacy policy says Muse may collect “information about you or others” from connected services.
Muse works through connectors that let it use services linked by its user. Meta’s help pages say email can be searched when connected, while the service does not download an entire inbox unless the user instructs it to. A calendar connector can send updates to Muse. Facebook, Instagram and Threads accounts may connect automatically when they share an Accounts Center with Muse. Access to Apple’s Mac Messages requires the user to grant macOS Full Disk Access and enable the connector.
Meta chief AI officer Alexandr Wang said the available connectors include Instagram direct messages and Messenger. The source material says Meta’s help pages do not document the precise scope of those two connectors. Meta’s privacy policy describes some connector data as shared proactively, while other access may occur when a task calls for it. Custom connectors are another route; Meta says it does not review those created by users.
Reports by WIRED and TIME described Muse instructions to keep pages about people in a user’s life and refresh information regularly. TIME reported that the agent’s files could include details about people named in messages and email, including relationships and disputes. Meta did not dispute TIME’s findings in the source material. These reports describe the agent’s instructions; they do not establish that every person mentioned is profiled in every use of Muse.
Infographic 01 · What Muse can read
Muse reads what its user can read. That includes you.
Connectors decide what the agent sees. The user switches them on. The people in that user’s chats and inbox are never asked.
Sources a Muse user can open to the agent
The agent reads what its user can read. Everyone who writes to that user is inside.
| Source | How it connects | Status |
|---|---|---|
| Email (Gmail, Outlook) | Opt-in connector. Searches when a task needs it; downloads the whole inbox only if told to. | Meta document |
| Calendar | Opt-in connector. Can push changes to Muse on its own (“proactive”). | Meta document |
| Facebook, Instagram, Threads | Connected automatically if the accounts share an Accounts Center with Muse. | Meta document |
| Instagram DMs, Messenger | Named as Muse-only connectors (“Instagram (incl dm’s) … and Messenger”). What they read is not documented. | executive post |
| Mac Messages (iMessage) | Only with macOS Full Disk Access plus the Messages connector, Meta says. One columnist says it synced anyway. | disputed case |
| A place to chat with Muse, not a connector. Test-build text mentions Muse access to end-to-end encrypted chats. | leaked code | |
| Custom connectors | Muse can build links to other services. Meta does not review them. | Meta document |
What the instructions say
Extracted by researchers; reported by WIRED and TIME reported
- “A page for every person in the user’s life”: family, partners, friends, colleagues, people followed
- Notes on how people met, shared interests, disputes, “tensions and alliances”
- People who never signed up are mapped by other users’ agents
- De-identified “lessons” shared across agents; Meta did not dispute TIME’s findings
What Meta says
Privacy policy, help pages, spokespeople Meta document
- Each user’s VM is isolated; no other user’s agent can reach it
- Conversations and VM data are not shared with Meta’s ad systems
- “Each person stays in control of their Muse and decides how much access it gets”
- Forgetting works “to the best of its ability”; deleted data may still be remembered
Sources: Meta Help Center: Connectors; Muse Privacy Policy; Alexandr Wang on X (8 Sep 2026); TIME (6 Oct 2026); Tom’s Guide on WIRED’s report (6 Oct 2026); TechCrunch (30 Sep 2026); Digit (16 Oct 2026). Status labels: Meta document = stated by Meta; executive post = Meta executive on X, not in help pages; disputed = claim and denial on record; leaked = unreleased code, may never ship.
Powered by Thorsten Meyer AI
Why Other People’s Messages Matter
The privacy issue extends beyond the person who installs the agent. If a Muse user connects accounts containing conversations, other people’s messages may become material the agent can search or use to build context. That means a person can be discussed in an AI system without creating an account or agreeing directly to its terms.
For readers, the practical distinction is between control of an agent and control of information about oneself. Meta says the Muse user chooses the agent’s access. A person on the other side of a conversation may have no equivalent switch in that user’s Muse settings. The scope depends on which connectors are enabled and what information they expose, details that are not fully described in the cited help pages.
The commercial stakes are also part of the story, but should be kept separate from the privacy findings. The source says Meta sold $25 billion in stock at $625 a share, alongside $12.5 billion of preferred stock, and Nvidia is investing $3 billion. Nvidia’s chips are to be deployed for Muse, according to the source. Those figures show a major financing and infrastructure relationship; they do not by themselves establish how Muse uses personal data or prove a particular business motive.
Popular right now

Shark Navigator Robot Vacuum with 30-Day Self-Emptying Base, White (AV2110S)

Breville BES870XL Barista Express Espresso Machine, Brushed Stainless Steel

De'Longhi Magnifica Evo Fully Automatic Espresso Machine with Manual Milk Frother
As an affiliate, we earn on qualifying purchases.
Muse’s Rollout and Connected Apps
Meta launched Muse in the United States on September 8, 2026, according to the supplied material. Japan followed on October 23, and Singapore was announced October 27 as its first market outside North America. Meta describes Muse as a cloud-based personal agent that can perform tasks such as handling email, booking tables and shopping. It says each user has a separate Muse Secure VM.
Muse can be used through its app, on a Mac or in a dedicated WhatsApp chat. The source distinguishes that chat from access to a person’s other WhatsApp conversations: opening Muse in WhatsApp does not, by itself, make it a reader of those other chats. TIME reported more than 5 million downloads within weeks; the source does not specify a precise measurement window beyond that description or provide an independent verification.
The reporting on people mentioned in Muse users’ data emerged in October. WIRED published findings on October 3 after security researcher Karan Joshi prompted Muse to return internal files, according to the supplied account. TIME reviewed the instructions separately three days later. The reports describe what the instructions direct the agent to do, while Meta’s public explanation emphasizes user control over access.
“Muse may still remember information that it learnt from”
— Meta privacy help page, as quoted in the supplied source
Open Questions on Data Retention
The supplied material does not provide a full technical account of what information each connector retrieves, how long derived information is kept, or which categories are used to create people’s profiles. Meta’s help pages reportedly explain some connector functions, but do not spell out the scope of Instagram DMs and Messenger access. The source also does not establish whether the reported hourly refresh instructions apply uniformly to all Muse users.
There is also a distinction between an agent’s internal instructions and its behavior in every real-world case. TIME and WIRED reported on those instructions, and Meta did not dispute TIME’s findings, but the material does not show how often profiles are created, what safeguards are applied in practice, or whether people mentioned in messages can inspect or correct information about themselves. Meta’s statement that users control access does not answer those questions for non-users.
Deletion is another unresolved area. The source says Meta’s help pages qualify Muse’s ability to forget information and warn that it may retain learned information after deletion. The full wording and exact retention mechanics are not included here, so it is not possible to say what is removed, what remains, or how long derived details persist.
What Users Can Change Now
Muse users can review and disconnect connectors, switch off model training, edit the agent’s memory file or reset it, according to the supplied source. Those actions affect the user’s own Muse environment; they do not give people mentioned in that user’s data direct control over the agent. Users should check which accounts are linked, particularly where services connect automatically through an Accounts Center, and review any custom connectors they have added.
For people who do not use Muse, the source points to WhatsApp Advanced Chat Privacy as one way to limit what can be retained or shared in other people’s chat archives. It does not present that setting as a complete block on information already available to a recipient or on other connected services. Meta’s public documentation and any further explanation of connector scope, deletion and profile retention will be important to follow as the Singapore rollout proceeds.
Meta has not, in the supplied material, announced a deadline for additional privacy controls or detailed answers to the reporting. The immediate next step for readers is to check their own chat and account privacy settings and ask contacts who use Muse which sources they have connected. Whether Meta will provide non-users with a way to review or challenge information held by someone else’s agent remains unknown.
Key Questions
Can Muse read messages from someone who has never signed up?
It may access information in services connected by a Muse user, including messages the user can access. Meta’s policy says Muse may collect information about other people. The supplied material does not establish that every message or every person is included.
Does Muse automatically read every WhatsApp chat?
No such access is confirmed in the source. It describes Muse in WhatsApp as a dedicated chat that a user opens, not as a reader of that user’s other WhatsApp conversations.
What controls can a Muse user use?
The supplied source says users can disconnect connectors, turn off model training, edit the memory file or reset Muse. Meta’s help material also cautions that forgetting information may not remove all information already learned.
Can non-users delete information held in another person’s Muse?
The source does not identify a direct control that lets a non-user manage another person’s agent. It describes some app-level privacy settings, including WhatsApp Advanced Chat Privacy, as ways to limit information in other people’s archives, not as a guarantee of deletion from Muse.
Source: ThorstenMeyerAI.com
