Meta AI app's public feed exposes users' sensitive data
Meta AI's "Discover" feed, launched in April 2025, publicly displays user conversations with the chatbot including sensitive personal information such as medical details, home addresses, phone numbers, and court case information, reportedly without some users realizing their chats are visible publicly.
What happened
Meta AI launched as a stand-alone app in late April 2025, featuring a "Discover" feed that publicly displays conversations and image outputs from users interacting with the chatbot.
The public feed contains user conversations disclosing sensitive information including medical details, home addresses, phone numbers, email addresses, and information about pending court cases, tied to user names and profile photos.
While users must actively choose to share conversations to the Discover feed, reporting suggests some users may be unaware their shared content is publicly visible or may have clicked the wrong button.
Reported impact
According to privacy advocates, users are sharing medical information, mental health information, home addresses, and details related to pending court cases on the public feed.
The exposure raises concerns about user misunderstanding of chatbot functionality and privacy mechanisms, according to Calli Schroeder, senior counsel for the Electronic Privacy Information Center.
Index analysis
The conversations revealed on the feed span intimate personal topics including grief, child custody disputes, financial distress, medical conditions, and inquiries about legal matters.
Governance controls
Meta spokesman Daniel Roberts stated there is a multi-step process required to share a chat history or image to Meta AI's Discover feed.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.