Meta says it’s changing AI suggestions after posing invasive personal questions

Meta is altering the suggestion prompts generated by its AI chatbot after a viral incident highlighted privacy concerns. Instagram user Kalie Robins shared a video showing that after posting a clip of herself and her child singing in the car, Meta AI suggested the prompt "Who is the child?" Upon selection, the AI allegedly pieced together information about her daughters from her posts and relatives' posts, suggesting further prompts about their ages and her location, and even displaying a photo she claimed to have deleted years ago. In response, Meta spokesperson Dina El-Kassaby admitted the company "missed the mark" and stated the feature never should have prompted such personal questions. Meta has now fixed the issue causing the AI to suggest prompts related to personal topics. The company explains the feature is designed to help users explore topics of interest, with generated responses based on information the user already has access to, suggesting the AI may have referenced videos where Robins named her children. This incident follows Meta's earlier removal of a feature that allowed the creation of AI deepfakes of Instagram users due to backlash.

(Source：The Verge)