Study reveals rising reliance of young people on AI chatbots for mental health support amid concerns over safety and privacy

A new study shows nearly one in five US young people are turning to AI chatbots for mental health advice, highlighting a growing trend that raises both benefits and serious safety concerns for parents, schools, and clinicians.

A new study suggests that artificial intelligence chatbots are already part of many young people’s mental health routines, with about 19% of US adolescents and young adults saying they have used them for advice on emotional or psychological concerns. The research, led by Ryan K. McBain of RAND and published in JAMA Network Open, points to a fast-growing behaviour that many parents, schools and clinicians may not yet fully see. Among those who used chatbots in this way, about 43% sought help at least monthly, and most said the advice was somewhat or very helpful.

That finding helps explain why chatbots have become appealing to teenagers in the first place. Unlike a search engine, a chatbot can carry on a conversation, adjust its tone and respond instantly, which can make it feel more personal and private than reading a webpage. For a young person who feels embarrassed, cannot easily reach a professional or worries about judgement, that combination of convenience and apparent confidentiality can be powerful.

But the same qualities also raise serious concerns. Chatbots do not have clinical judgement, and fluent answers can sound more certain than they really are. They may miss abuse, coercion, self-harm risk or other warning signs, especially when distress is expressed indirectly. Researchers say that is why simple counts of who is using these systems matter: they help establish the scale of a behaviour that could bring both benefits and harms.

The privacy issue is just as important. Teenagers may reveal names, symptoms, family conflict or other sensitive details without realising how those conversations may be stored, reviewed or used. The study’s findings suggest many young people are not telling anyone they are using chatbots for mental health advice, which means parents and carers may have little idea where support-seeking is taking place. That leaves schools, doctors and platform makers facing the same challenge: how to preserve easy access to digital tools without letting them stand in for trained human help when a child is at risk.