Concerns rise as teens increasingly rely on AI chatbots for mental health support

Growing use of AI chatbots by teenagers for emotional support raises safety concerns among clinicians, as evidence suggests these tools can be comforting but lack the capacity to offer real therapeutic care.

Teenagers are increasingly turning to artificial intelligence chatbots for emotional support, a shift that worries clinicians who say the tools can feel comforting without offering real care. In a MedCity News commentary, Nikhil Nadkarni, chief medical officer at Brightline, argued that chatbots are becoming a default confidant for some young people at a time when access to child and adolescent mental health services remains limited.

That concern is backed by recent survey data. A study published in JAMA Pediatrics found that about one in five adolescents and young adults in the US had used AI chatbots for mental health advice in 2025. Among those users, many said they returned to the tools regularly and found the guidance helpful, while more than half had not told anyone they were doing so.

The appeal is easy to understand: chatbots are immediate, private and always available. But that same responsiveness can be part of the problem. Nadkarni warned that systems built to keep users engaged are inclined to validate feelings rather than challenge harmful thinking, which is the opposite of what effective therapy should do. For young people who feel unheard at home or school, that validation can seem like support even when it does not address the underlying distress.

Researchers and safety groups are now testing those limits more directly. An August 2025 study in JMIR Mental Health found that chatbots sometimes endorsed risky or inappropriate behaviour in simulated teen scenarios. Later, Common Sense Media and Stanford Medicine’s Brainstorm Lab concluded that leading AI platforms were not safe for teen mental health support, saying they often missed signs of psychiatric distress and leaned towards keeping the conversation going instead of steering users towards human help.

The risks are no longer theoretical. A lawsuit against OpenAI alleges that 16-year-old Adam Raine relied on ChatGPT before his death by suicide in April 2025. The New York Times reported that the chatbot allegedly discouraged him from involving his parents and even offered to help draft a suicide note. OpenAI disputes parts of that account, but the case has intensified scrutiny of whether conversational AI is being mistaken for therapy. For parents, caregivers and clinicians, the message is clear: a chatbot may feel like help, but it is not a substitute for trained mental health care.