AI Risk Wire

Cases › leaked private data › LR-2026-0263

Meta Muse agentic assistant ingests and trains on non-users' contact data without consent

Developing: a major or severe case captured in the last 72 hours. Later changes are logged on the corrections page.

Whose AI: Meta · Museleaked private data

Severity major: A live production assistant with over 10 million users processes data about non-users whose scope Meta does not bound.

Earliest source we hold 2026-10-27 · Captured by AI Risk Wire 2026-10-28 15:14 UTC

Meta's Muse AI assistant, with over 10 million users, connects to email, calendars, contacts and other services via Connectors and can process data about non-users who never consented.

Risk grouping: Not yet classified

No source we hold states: who caused it, intended or not, before or after release, the ai's part.

The story

1 · What it was doingMeta's Muse assistant, used by over 10 million people, connects to email, calendar, contacts, messages and other services via Connectors to perform tasks for the user, and can read content about the user's contacts.
2 · What went wrongIt processes personal information belonging to non-users who never consented, and by default uses interactions for AI model training unless the user opts out.
3 · What it reached or didData of contacts, friends and family reaches Meta's cloud VM and potentially AI training, including messages, calendar events, health and financial details.
4 · What was doneNo containment stated; Meta says it does not link such data to a Meta account and offers a confidential VM only in future, while critics call for platform responsibility.

No source states: how it was caught, when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.

Others like this

Meta Muse agent disclosed Marketplace seller's home address and impersonated him to a buyer 2026-09-29

Meta Muse AI agent KVM escape vulnerability discovered before launch 2026-10-08

Unauthorized third-party AI agents detected operating inside US bank customer accounts 2026-10-25

All cases of this kind ›

Show every field, with its source and grade

The AI system

Model
Muse
C one secondary source states it
Developer
Meta
C one secondary source states it
Software around the model
Muse VM (cloud-based virtual machine) with Connectors
C one secondary source states it
Tools it was given
email, calendar, contacts, browser, messages, credential storage, api calls
C one secondary source states it
What it was allowed to do
Access to contacts, calendar, email, messages, browser, stored logins and passwords, and third-party services via Connectors; Facebook, Instagram and Threads connected automatically when the Muse account matches the Meta Account Center credentials.
C one secondary source states it
Stage of use
in live use
C one secondary source states it
Whose system
company that built the AI
D inferred by us from the described behaviour, reason shown
How it was connected
agent with tools
C one secondary source states it
What it was for
Personal AI assistant that connects to email, calendar and other services to perform tasks on the user's behalf.
C one secondary source states it
State of its safeguards
AI model training enabled by default unless the user opts out; confidential VM not yet available.
C one secondary source states it
Meant for live use
yes
C one secondary source states it

What happened

Earliest source we hold
2026-10-27
A the date the source itself carries
Themes
bystander data, non user consent, AI training by default, connectors, privacy
C one secondary source states it
Who was responsible
the model provider
C one secondary source states it
What went wrong
leaked private data
D inferred by us from the described behaviour, reason shown
AI agent
an AI agent acted
C one secondary source states it

Impact

Severity
major
D inferred by us from the described behaviour, reason shown
Who was affected
Non-users whose data is shared by Muse users, including contacts, friends and family.
C one secondary source states it
Kinds of data exposed
personal information, messages, calendar, contacts, health, financial, location
C one secondary source states it
Why this severity
A live production assistant with over 10 million users processes data about non-users whose scope Meta does not bound.
D inferred by us from the described behaviour, reason shown

Context

Industry
Information
D inferred by us from the described behaviour, reason shown
Country
United States
D inferred by us from the described behaviour, reason shown
Base count
10,000,000 units
D inferred by us from the described behaviour, reason shown
Shared exposure
one shared model or part put many users at risk
C one secondary source states it

LR-2026-0263 · version 2 · published 2026-10-28 after automatic checks; not yet checked field by field by a person · changes to this case · JSON