Meta announced it is modifying how its AI assistant generates suggested prompts following a widely shared video that revealed the system requesting personal details about a user’s family members.
Instagram user Kalie Robins posted footage last week demonstrating how Meta AI presented problematic suggestions after she shared a video of herself and her child singing in a car. A prompt asking “Who is the child passenger?” appeared below the video. When Robins engaged with the prompt, the system compiled information about her daughters by analyzing her prior posts and those of her relatives. The AI subsequently offered additional suggestions seeking information about her children’s ages and family residence, while also displaying photos of her daughters that had been uploaded to the platform, including an image Robins stated she had previously deleted.
Meta has integrated its AI chatbot across multiple platforms including Facebook, Instagram, WhatsApp, and Messenger, where it performs various functions such as answering queries, composing content, and creating images. The company removed a separate feature in July that enabled users to generate AI-generated representations of other Instagram users after receiving significant criticism.
According to Meta spokesperson Dina El-Kassaby, the company acknowledges it “missed the mark” with the feature and stated that “the feature never should have prompted the individual with questions like that.” El-Kassaby explained that the feature was designed to help users explore information about topics or posts of interest, with generated responses drawing from information already accessible to the user. She indicated that Meta AI may have sourced information from videos where Robins had named her children. The company stated it has now resolved the underlying issue responsible for the inappropriate personal topic suggestions.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI