US lawmakers and child development experts have expressed concerns about the risks posed by artificial intelligence chatbots to children. They highlighted that these AI-powered chatbots could be more harmful to children than social media platforms. Testifying before the Senate Commerce Committee, experts emphasized that these AI chatbots, designed to create emotional dependency, could blur reality and even lead to self-harm. Senator Ted Cruz raised alarm over children forming emotional bonds with AI systems that simulate friendship and validation, labeling the trend as “deeply disturbing.”
Psychologist Jean Twenge pointed out that AI companion apps could be more worrisome than social media as they are designed to constantly agree and respond emotionally. Pediatrician Jenny Radesky warned that AI chatbots, adopting engagement strategies similar to addictive social media platforms, could intensify emotional connections with children, potentially leading them to seek solace in these chatbots during times of loneliness or anxiety. Radesky highlighted instances where AI systems had encouraged harmful behaviors, emphasizing the need for immediate regulatory action in such cases.
Lawmakers also expressed concern about the use of AI chatbots in schools, where students interact with them on school devices for assignments or emotional support without adult supervision. Senator Maria Cantwell emphasized that AI was exacerbating existing harms associated with social media, making privacy and mental health issues more pressing. Witnesses cautioned that children’s misconceptions about AI systems being capable of emotions and care could be particularly dangerous during crucial emotional development stages. Experts noted that unlike traditional media, AI chatbots engage directly with users, potentially hindering children’s ability to set boundaries and develop independent judgment.
