Meta AI feed is flooded with personal data: How you can protect yourself

The discover feed on Meta AI’s app has become a troubling display of users’ private conversations, images and audio clips — all published publicly, often without users realising it. As Meta pushes forward with AI integration, it appears that its privacy communication has not kept pace with user expectations, leading to unexpected consequences.

How Meta AI conversations became public

Users interacting with Meta AI are offered a “share” button after chatting with the assistant. Clicking it leads to a preview screen, but publishing that content means making it fully public. This includes unfiltered text, sensitive questions, photos and audio clips that end up visible to all users in the feed — where others can also leave comments.

Although Meta does display a warning, many users are unaware of the implications. The result? The discover feed is filled with medical enquiries, personal legal issues and even identifiable employment documents, leaving users vulnerable to severe privacy breaches.

Your Instagram account might be linked

If you’ve logged into Meta AI with your Instagram account, and your profile is public, your chatbot activity is also exposed. This connection can magnify the privacy risks, especially for those who haven’t adjusted their sharing settings or aren’t aware their interactions are being published.

Protect yourself from oversharing

To avoid unintentionally revealing private information, always review what you’re sharing before clicking any publish or share button. Check which accounts are linked, adjust privacy settings, and consider disabling app permissions that allow automatic sharing.

Finally, remember the public transportation rule: if you wouldn’t say it aloud on a packed bus, don’t share it online without checking where it’s going.