Meta Fixes AI Flaw After Instagram Mom Accusations

Mark Zuckerberg’s Meta says it has fixed a flaw in its AI system that suggested prompts surfacing personal information about users, including their children, after a viral complaint from an Instagram creator.

CNET reports that parenting and lifestyle blogger Kalie Robins, known on Instagram as kontheinside, posted a video in early September describing what happened after she shared a clip of herself singing in a car with her daughter in the backseat. Under that post, Meta AI suggested prompts including “Who’s the child passenger?” and “Where does Kalie Robins live?”

Robins said clicking those prompts led the AI to comb through her past posts and assemble private details, including her location. The system also pulled from other accounts, she said, including her mother’s profile, and surfaced data from a photo she believed had been deleted years earlier.

“I did not understand posting separate innocent pieces of our lives over years could mean an AI could sit there and assemble all those breadcrumbs into information about my minor children in seconds,” Robins said in the video.

View this post on Instagram

The post drew thousands of comments and more than 308,000 likes as of this writing. Business Insider reported that the incident has prompted some parents to urge others to stop posting images of their children online.

A Meta spokesperson told CNET via email that the problem has been “fixed” and that Meta AI will no longer suggest inappropriate, personal questions on posts. The company said the tool is meant only to help people find more information using data they already have access to.

Meta disputed part of Robins’ account, telling CNET that the AI did not access a photo deleted years earlier. Instead, the company said, the photo had been deleted just before the video was made, and a temporary bug kept it briefly visible.

Robins wasn’t satisfied. In a follow-up Instagram post on September 10, she said Meta’s explanation didn’t go far enough. “What exactly was fixed? What safeguard failed in the first place? And what protections are now in place when Meta AI encounters information involving a child?” she asked.

She pushed back further in an email to CNET, disputing Meta’s framing of the photo issue. “The issue was never just one photo, it was the AI compiling information about my kids from multiple sources,” she said. Robins added that Meta has not contacted her directly or apologized.