As much as I’ve enjoyed using Meta’s Ray-Bans, I haven’t been a very big fan of the switch/rebrand from the Meta View app, which was a fairly straightforward companion to the smart glasses. Now, we’ve got the Meta AI app, a very not-straightforward half-glasses companion that really, really tries to get you to interact with—what else—AI. The list of reasons why I don’t like the app transition is long, but there’s always room for more grievances in my book, and unfortunately for Meta (and for us), that list just got a little bit longer.

Wild things are happening on Meta’s AI app.

The feed is almost entirely boomers who seem to have no idea their conversations with the chatbot are posted publicly.

They get pretty personal (see second pic, which I anonymized). pic.twitter.com/0Hoff1psPU

— Justine Moore (@venturetwins) June 11, 2025

There were a lot of tweaks when Meta crossed over from the Meta View app to the Meta AI app back in late April, and it seems not all of them have been registered by the people using it. Arguably one of the biggest shifts, as you can see from the tweet above, is the addition of a “Discover” feed, which in this case means that you can see publicly, by default, what kinds of prompts people are funneling into Meta’s ChatGPT competitor. That might be fine if those people knew that what they were asking Meta AI would be displayed in a public feed that’s prominently featured in the app, but based on the prompts highlighted by one tech investor, Justine Moore, on X, it doesn’t really look like people do know that, and it’s bad, folks. Very bad.

I spent an hour browsing the app, and saw:
-Medical and tax records
-Private details on court cases
-Draft apology letters for crimes
-Home addresses
-Confessions of affairs
…and much more!

Not going to post any of those – but here’s my favorite so far pic.twitter.com/9KqGeLB5UN

— Justine Moore (@venturetwins) June 12, 2025

As Moore notes, users are throwing all sorts of prompts into Meta AI without knowing that they’re being displayed publicly, including sensitive medical and tax documents, addresses, and deeply personal information—including, but not limited to—confessions of affairs, crimes, and court cases. The list, unfortunately, goes on. I took a short stroll through the Meta AI app for myself just to verify that this was seemingly still happening as of writing this post, and I regret to inform you all that the pain train seems to be rolling onward. In my exploration of the app, I found seemingly confidential prompts addressing doubts/issues with significant others, including one woman questioning whether her male partner is truly a feminist. I also uncovered a self-identified 66-year-old man asking where he can find women who are interested in “older men,” and just a few hours later, inquiring about transgender women in Thailand.

I can’t say for sure, but I am going to guess that neither of these prompts was meant for public consumption. I mean, hey, different strokes for different folks, but typically when I’m seeking dating advice for having doubts about my relationship, I prefer it to be between me and a therapist or close friend. Gizmodo has reached out to Meta about whether they’re aware of the problem and will update this post with a response if and when we receive one. For now, it’s advisable, if you’re going to use the Meta AI app, to go to your settings (or your parents’ settings) and change the default to stop posting publicly. To do that, pull open the Meta AI app and:

• Tap your profile icon at the top right.
• Tap “Data & Privacy” under “App settings.”
• Tap “Manage your information.”
• Then, tap “Make all your prompts visible to only you.”
• If you’ve already posted publicly and want to remove those posts, you can also tap “Delete all prompts.”

I’ve seen a lot of bad app design in my day, but I’ll be honest, this is among the worst. In fact, it’s evocative of a couple of things, including when Facebook released a search bar back in the day that was misconstrued for the post bar by some, causing users to type and enter what they thought was a private search into the post field. There’s also a hint of Venmo here when users were unaware that their payments were being cataloged publicly. As you might imagine, those public payments led to some unsavory details being aired. For now, I’d say it’s probably best to steer clear of using Meta AI for anything sensitive because you might get a whole lot more publicity than you bargained for.