• Source:JND
Add as a preferred

Imagine waking up to find your personal conversations with an AI assistant published for the world to see. That’s exactly what’s happening with Meta’s new stand-alone Meta AI app, and it’s raising serious privacy concerns.

Users Are Unknowingly Sharing Conversations Publicly

The Meta AI app features a button labelled “Share” that allows users to post excerpts of their conversations with the AI. It now seems that many users are oblivious to the fact that these excerpts can be viewed by anyone. Users can post sharable text conversations and audio clips, along with personal pictures, which subsequently get posted to a feed visible to all.  

One particularly surreal example: An audio recording of a man with a Southern accent asking, “Hey, Meta, why do some farts stink more than other farts?”. That’s just the tip of the iceberg.

ALSO READ: Midjourney In Legal Trouble As Disney, Universal File Copyright Lawsuit

The Real Risk: Exposing Personal and Legal Information

While flatulence queries may be humorous, other posts are deeply concerning:

  • Users discussing tax evasion
  • People asking if family members might be arrested
  • Requests for help writing character references for individuals facing criminal charges, with full names included
  • Posts containing home addresses, court records, and other sensitive personal data

Cybersecurity expert Rachel Tobac has even flagged examples of dangerous oversharing, warning that this could lead to identity theft or real-world consequences.

No Clear Privacy Notices from Meta

What makes this issue worse is Meta’s lack of transparency. Users have no clear indication of:

  • What privacy settings are enabled when posting?
  • Where the content is being shared?
  • Who can view their shared conversations?

If you’ve signed in with a public Instagram account, chances are your Meta AI conversations are public too — including those asking how to meet “big booty women.”

A Design Flaw or a Feature Nobody Wanted?

Basic UX foresight could have avoided this entire debacle. It’s hard to fathom why anyone at Meta would think users would want a social feed of AI queries. History has offered us many lessons of the dangers of such thinking. Remember AOL’s PR disaster after releasing anonymized user search logs in 2006?  

There’s a reason Google never turned its search engine into a public forum — some things are better left unsaid.

ALSO READ: Sam Altman’s AI Optimism Can’t Hide OpenAI’s Environmental Reality

A Modest Launch, But a Growing Mess

As of now, the Meta AI app has been downloaded 6.5 million times since April 29, according to app intelligence firm Appfigures. While these numbers aren’t terrible, for a company of Metas size and resources, this is underwhelming.

Yet, the app’s public feed is already descending into trolling chaos. Posts include:

  • Job seekers sharing résumés while asking for cybersecurity roles.
  • Users with Pepe the Frog avatars requesting tutorials on how to make a water bottle bong

Conclusion: Viral Growth by Public Embarrassment

If increasing user engagement through controversy is what Meta was trying to achieve, they are bluntly succeeding – but at a great price. The AI app Meta intended to showcase as a demonstration of their multi-billion dollar investment in AI is quickly becoming a privacy minefield and a viral trainwreck.

Until privacy measures are put in place, Meta users might want to reconsider having conversations with Meta AI — or at least reconsider pushing that share button.