Meta AI Suggested Invasive Prompts About Users' Children

Meta addresses AI prompt controversy

Meta has disabled a feature in its Meta AI software that generated invasive question prompts about users' personal lives, including information about their children. The issue came to light after a viral Instagram video demonstrated how the AI tool suggested deeply personal queries and assembled private details from years of social media activity.

Parenting blogger Kalie Robins, who posts as kontheinside, documented the problem in a September 2 video that has since garnered more than 308,000 likes. After posting a video of herself singing in a car with her daughter, Robins noticed Meta AI had generated prompts beneath the post asking "Who's the child passenger?" and "Where does Kalie Robins live?"

When she clicked these automated questions, the AI pulled information from multiple sources across her posting history. The tool compiled data from her previous posts, accessed her mother's account, and even referenced what Robins claimed was a photo deleted years earlier. The AI assembled these fragments into a detailed profile that included her location and information about her minor children.

Why it matters

This incident highlights the privacy risks inherent in AI systems that can aggregate and synthesize information from disparate sources. For parents who have shared years of family content online, the ability of AI to instantly compile scattered details into comprehensive profiles raises urgent questions about consent, data retention, and child safety — especially as Meta and other platforms roll out increasingly sophisticated AI features.

Company response and lingering questions

A Meta spokesperson told CNET the company has fixed the issue and that Meta AI will no longer suggest inappropriate personal questions on user posts. Meta stated the feature was intended only to help people find information they already have access to, and acknowledged it missed the mark.

Meta also disputed Robins' claim about the deleted photo, telling CNET the image was removed shortly before she made the video and remained briefly visible due to a temporary bug.

Robins responded that Meta's statement didn't address her core concerns. In a September 10 follow-up post, she asked what specific safeguard failed, what exactly was fixed, and what protections now exist when Meta AI encounters information involving children. She told CNET that Meta has not contacted her directly or issued an apology.

"The issue was never just one photo — it was the AI compiling information about my kids from multiple sources," Robins said.

Broader privacy concerns

The controversy adds to Meta's ongoing struggles with child safety and privacy. The company has faced multibillion-dollar settlements and court judgments this year related to how its technology affects children, including issues of addiction and exploitation.

Meta is simultaneously expanding its AI capabilities through new products like Muse AI agents, marketed as personalized AI assistants, while facing criticism over facial recognition technology in its smart glasses.

Some parents are now urging others to stop posting images of their children online entirely, according to reporting by Business Insider. Privacy advocate Cathy Pedrayes, who frequently posts about scams and dark-web trends, called the incident "creepy."

The story was first widely publicized by Futurism and reported in detail by CNET.