You May Not Be Aware Of This But Your Meta AI Searches Are Public Knowledge

If you’ve been using Meta’s new AI tool to solve maths homework, write poetry, or ask deep personal questions, here’s a plot twist you probably didn’t see coming: some of those prompts are being posted publicly – and your username might be attached.

Meta AI, the tech giant’s answer to ChatGPT, includes a public “Discover” feed showcasing prompts submitted by users and the AI’s responses. While Meta insists chats are private by default, it turns out a lot of users are still accidentally making their conversations visible to the world – including some that are, well, pretty personal. 

MORE: WhatsApp is making 3 major changes and we are obsessed

DISCOVER: Should you be hiding your children’s faces online?

Image @healinfairy via Pinterest

Think: people uploading photos of school test questions for help, exploring questions about gender identity, or requesting “scantily clad” anthropomorphic animal characters. One prompt even asked for an image of an animated character lounging in their underwear – and it was publicly visible along with the user’s Instagram profile pic and handle.

So how’s this happening? Meta says that before any post is shared, users get a warning: “Prompts you post are public and visible to everyone… Avoid sharing personal or sensitive information.” But critics argue that the message isn’t exactly crystal clear – especially for users who aren’t aware that their AI chat can end up in what looks a lot like a social media feed.

MORE: Will Apple’s iOS 18 Make It Easier to Cheat in Relationships?

DISCOVER: The best personal safety gadgets to carry with you

“If a user’s expectations about how a tool functions don’t match reality, you’ve got yourself a huge user experience and security problem,” said Rachel Tobac, CEO of cybersecurity firm Social Proof Security. She added that because the Discover feed looks like just another part of social media, “users are inadvertently posting sensitive info to a public feed with their identity linked.”

And that’s the crux of it. While Meta claims, “You’re in control: nothing is shared to your feed unless you choose to post it,” it seems many people don’t realise what exactly they’re “choosing.”

MORE: 21 Features We Loved from Apple’s WWDC24

DISCOVER: TikTok’s Ticking Time Bomb: Will the US Ban The Viral App? – Here’s What You Need To Know

Meta AI is currently live in the UK via browser, and in the US via app – and it’s integrated into Facebook, Instagram, and WhatsApp. So yes, your AI activity could be just a few clicks away from your actual digital identity.

For those keeping score: public-by-default content, auto-linked usernames, and generative AI… what could go wrong?

If you’re using Meta AI, now’s a good time to double-check your privacy settings. Because no one wants their experimental furry fan art or exam cheats floating around the internet with their profile pic attached.