Meta AI Was Building Profiles of a Mother's Children From Old Family Posts
A viral video showed Meta's built-in chatbot compiling details about a woman's young daughters, including a photo she thought she'd deleted years ago.
Key points
- Meta AI surfaced a deleted photo of a user's daughters and offered prompts asking about the children's ages and the family's home location.
- Meta spokesperson Dina El-Kassaby admitted the company "missed the mark" and said the feature "never should have prompted the individual with questions like that."
- Meta has since changed the suggested prompts its AI chatbot generates for personal topics.
- Meta pulled a separate feature in July that let users create AI deepfakes of other Instagram users after public backlash.
Last week, Instagram user Kalie Robins posted a video describing what happened after she cross-posted a clip of herself and her child singing in the car. Beneath the video on Facebook, Meta AI, the chatbot built into Facebook, Instagram, WhatsApp and Messenger, displayed a suggestion: "Who is the child passenger?"
She tapped it. What came back wasn't a simple answer.
Meta AI pulled details about her daughters from her own past posts and from posts her relatives had made, then offered follow-up questions about the children's ages and where the family lives. It also surfaced photos of her daughters, including one Robins says she deleted years ago.
Meta confirmed the incident to The Verge and called it a mistake. "We created this feature to help people get more information about topics or posts that interest them," El-Kassaby said. She added that the AI may have drawn on videos in which Robins had named her children, and that any information it surfaced came from content the user already had access to.
That last point is technically accurate but sidesteps the real concern. No single piece of information was secret. What was alarming was that the AI assembled scattered family details into a ready-made profile, then volunteered to add more, all without Robins asking for any of it. Our earlier story on 10 September found that Meta's Muse AI agent can pull personal details from social accounts that even Instagram won't surface directly, which puts this incident in a pattern, not an isolated slip.
Why does this matter for ordinary users?
Meta AI isn't a separate app you choose to open. It's woven into platforms most people already use daily, which means it can act on your posts, and on posts other people make about you, without you realising it.
Most people don't think of a comment a cousin left under a birthday photo as data an AI could stitch together with other posts to sketch a picture of their family. Meta's own system just showed that it can.
Meta says it's fixed the prompts that triggered this kind of profiling. The company didn't say whether the underlying behaviour has changed, or only the suggested questions that appear on screen. That's the thing to watch.
Should you worry about your own posts?
If you use Facebook or Instagram, Meta AI may already be suggesting prompts on your posts or your family members' posts. You don't have to tap them. Treat any AI-generated suggestion about a person, especially a child, the way you'd treat a stranger asking the same thing out loud.
Check your privacy settings on both platforms and think about what you, and people close to you, have made visible. Old posts you've forgotten are still there, apparently still findable. If you see AI-generated prompts asking for personal details about someone shown in a photo or video, report them rather than engaging with them. Meta's fix may not catch every variation.
