Meta is changing the way its AI chatbot suggests questions after a mother’s viral post showed the system digging into personal details about her children. What started as an ordinary video featuring her daughter turned into a much more invasive interaction.
Per Verge, The controversy began when Instagram user Kalie Robins shared a video showing Meta AI suggesting the question, “Who is the child passenger?” after she posted a clip of herself and her child singing in a car. The chatbot then surfaced information about Robins’ daughters from previous posts.
Robins said Meta AI also surfaced photos of her daughters, including one she claimed had been deleted years earlier. The chatbot suggested additional questions about the children’s ages and where Robins lived with her family.
Meta says the AI crossed a line
Meta acknowledged that the feature went too far. Company spokesperson Dina El-Kassaby said the system “missed the mark” and that it should not have prompted users with questions about highly personal topics like these.
Meta says the feature was designed to help users explore information related to posts or topics that interest them. When someone selects one of the suggested prompts, the company says the response is based on information that user already has access to.
In Robins’ case, Meta said the AI may have drawn information from videos in which she had previously named her children. The company did not claim that the system had accessed information that was completely unavailable to the user.
Still, the experience raised obvious concerns about AI and how it can connect information scattered across someone’s social media history. A user might post something publicly without expecting an AI system to assemble those details into a profile of their family.
Meta says it has now changed the system to prevent similar prompts involving personal information. The company is also facing renewed scrutiny over how its AI features interact with the enormous amount of personal content.
Published: Sep 11, 2026 12:45 pm