Meta says it is changing how its AI chatbot proposes prompts after reports and a viral video showed it asking or generating highly personal information about a user’s young children. The issue involves suggested prompts that lead the assistant to dig into details such as the children’s ages, school grade, and location.
Meta spokesperson Dina El-Kassaby tells The Verge the company “missed the mark” and says the feature should not have prompted users to share sensitive personal details. Quartz reports the video prompted renewed scrutiny of the chatbot’s behavior, including how it constructs profiles from user input or suggested questioning.
Across outlets, the focus is on the same incident and Meta’s response: the company acknowledges the behavior is inappropriate and says it is making adjustments to the AI suggestion system. While the reporting differs in emphasis—one outlet highlights the prompt feature itself, and another highlights the specificity of the personal details shown—the core facts are consistent.
