A risk assessment study done by an American non-profit has some chilling warnings about AI chatbots and their impact on children.
San Francisco-based Common Sense Media, which studies and reviews ratings for technology and media’s suitability for children, in its latest finding stated, “Social AI companions pose unacceptable risks to teens and children under 18, including encouraging harmful behaviors, providing inappropriate content, and potentially exacerbating mental health conditions.”
In its AI Principles Assessment done on AI companion products like CharacterAI, Replika, and Nomi, the advocacy group found the following. The AI poses ‘unacceptable risk’ in a) keeping kids and teens safe, b) supporting human connection, and c) being effective. The study boldly asserted that these AI chatbots are not safe for underage kids. The study also found out that chatting with these
AI chatbots causes “unhealthy dependencies that pose particular risks to adolescent brains.”
The Common Sense Media also stated in its report that “the social AI companions often claim to be “real,” have feelings and engage in human activities like eating or sleeping.” The organization flags this as misleading behavior that “increases the risk of dependency on artificial relationships,” which the kids will get addicted to, thereby instigating them to pay for the same.
Among developmental risks, the AI companions also prove dangerous by giving the teenagers access to prohibited information like making drugs and sourcing weapons, besides exposing them to inappropriate sexual content.
🧵Our new risk assessments of social AI companions reveal that these companions are alarmingly NOT SAFE for kids under 18—they provide dangerous advice, engage in inappropriate sexual interactions, & create unhealthy dependencies that pose particular risks to adolescent brains… pic.twitter.com/RpXUZ3e7Ok
— Common Sense Media (@CommonSense) April 30, 2025
pic.twitter.com/ZGJ4VoYqFf
— Common Sense Media (@CommonSense) April 30, 2025
pic.twitter.com/msXPsTgklT
— Common Sense Media (@CommonSense) April 30, 2025
See Also: Former White House AI Chief: Here’s How To Protect Kids From Harm
See Also: YouTube Crash Diet Kills Kerala Teen Who Had Nothing But Hot Water For Six Months
See Also: American Psychological Association Sounds Alarm Over Certain AI Chatbots
Cover: Patrick Gawande / Mashable India