Your Questions in the Meta AI App May Be Published Publicly
I never thought I’d download Meta AI on my iPhone. After all, people have been angry for over a year that you can’t turn off Meta AI on Facebook and Instagram : why would you need a separate app for that?
Then I saw headlines from TechCrunch , Wired , and Business Insider , among many others, that were highly critical of the app’s approach to privacy and security. That’s because Meta AI isn’t Meta’s answer to ChatGPT, Gemini, or Claude. Instead, it’s part chatbot, part social media platform, where your queries and questions can be shared with the rest of the Meta AI community.
To be clear, your Meta AI interactions are not published by default. You need to choose to publish your requests in the social aspect of the app. If you choose to do so, your requests will be published in a public Discover feed that anyone who has the app can see. This, of course, invites users who want to share their AI creations, as well as trolls who want to spam the feed with stupid or offensive requests and generations. But what’s even more disturbing about the feed is that it hosts messages from users who clearly didn’t realize they were posting their chats publicly.
Meta AI’s Discover Feed Is a Spectacle to Behold
My first impression when scrolling through this feed was that most users seem to be genuinely interested in the social aspect of the app. Some posts seem to be geared toward a public audience, with users commenting on them as if they were commenting on Instagram or Facebook posts. And, at its core, most of these posts are pretty innocuous: a Maltese swimming in a pool; killer clowns from outer space; Nartuo and Deku fighting an epic battle in the multiverse; and a lot of anthropomorphic animals… so many anthropomorphic animals.
Even some personal posts are audience-friendly: I saw a screenshot of someone’s sleep stats, supposedly tracked from an Apple Watch, and the user asked Meta AI to analyze the stats and let me know what it thought. The user then responded to comments, including one who said “the audit looks good” and one who wondered if the user sleeps with “these” — “these” presumably meaning smartwatches? Why you’d want to start a public discussion about your sleep habits with strangers is beyond me, but to each their own.
But every now and then, you come across something that was clearly meant to remain private — at least between the user and Meta — and is now visible to me. I saw someone post a really close-up selfie asking for a pretty but realistic makeover. They seemed unhappy with the digital, carnival-like makeover the AI was using, because they asked again for it to look realistic. (It wasn’t.) Someone asked for “a skinny cute anthropomorphic lion girl in pink striped socks dancing on a pole at a club.” (I can report that Meta AI actually animated that.) Then the same person posted an image of themselves with two women kissing in the background (these women were added by Meta AI in a separate request) and asked Meta to keep adding more and more women to the background. The only comment I saw was, “they’re posted publicly, my boyfriend.”
It’s not just that these posts are publicly available: They’re tied to your Meta account, which anyone can open to see your entire posting history. One user who asked Meta to generate “dirty bikinis and passionate kisses” also asked Meta AI what to do about a few red bumps on her inner thigh. Sorry to put you through this, but also, how do I and the rest of the Meta AI app know about this? I’m going to go out on a limb and guess you didn’t mean to share it — unless you’re just messing with us.
Meta-AI Is a Privacy and Security Nightmare
Meta has a poor reputation for privacy in general. (It is, after all, the company that allowed Cambridge Analytica to leak 87 million accounts .) But the app is a privacy and security nightmare .
If you look at the iOS App Store’s app privacy report, you’ll see that Meta AI collects a ton of your data, including health and fitness information, financial information, contact information, browsing history, usage data, your location, contacts, search history, sensitive information, and “environment,” which I didn’t even know was a metric. ( According to Apple , this means “environmental scanning,” including “grid, plane, scene classification, and/or image detection of the user’s surroundings.”)
But dig into the settings and you’ll find some terrible defaults. For one, Meta will suggest your publicly posted suggestions to you on other platforms like Facebook and Instagram. (You can turn this off in the Meta AI app settings by tapping your profile icon, then Data & Privacy > Suggest your suggestions to other apps .) Meta AI also supports “background conversations,” which essentially listen to you when you exit the app or put your phone to sleep, in case you want to continue chatting with Meta AI at any time. No thanks. (You can turn this off in Settings > Data & Privacy > Voice .)
There are some serious security concerns here, too. One person shared a photo of their computer (taken with a camera, not a screenshot, mind you) that had a warning that their Facebook account would be disabled in 169 days. As it turns out, Facebook had banned their account, but the user had appealed the ban and didn’t know what to do next. The user gave their full name and asked if the bot could “talk to this AI about my appeal?” It then, without prompting, told me what kind of business they were in, which was enough information for me to find their LinkedIn and Instagram accounts. I’m pretty sure this person had no idea that their requests to Meta AI had been posted publicly, and I’m guessing that if they had, they wouldn’t have shared those details so casually — or communicated with Meta AI in the first place.
Too many people seem unaware that the posts they share with Meta AI about themselves, friends, and family, or the deeply sensitive questions they entrust to the bot, are now available to the community of users scrolling through the Discover tab. Of course, there’s always the possibility that any of these posts are clever trolls, especially given the recent media attention Meta AI has received. But something tells me that the person asking for help getting their account locked is a real person who had no idea their desperate conversation with Meta’s product would end up in a public feed, let alone in this article.
Make all your Meta AI public posts private
If you’ve been posting messages in the Meta AI app without realizing it, or you simply regret all your public messages, you can make them all private in one fell swoop. To do this, go to Settings > Data & Privacy > Information Management . Tap “Make all public requests visible only to you,” then “Apply to all” to make all messages private. Or, you can tap “Delete all requests,” then “Delete all” to get rid of them forever.