RAND researchers found that 19% of people ages 12 to 21 used AI chatbots for advice when they felt sad, angry, nervous or stressed, up from 13% earlier in 2025, with findings published Monday in JAMA Pediatrics. Clinically, the study suggests AI is becoming a parallel mental health touchpoint for youth at roughly the scale of professional therapy use — often without disclosure, safety standards or crisis-ready oversight.

The Move

• Nearly 1 in 5 adolescents and young adults reported using tools like ChatGPT, Google Gemini or Character.AI for mental health advice.

• Most users said the advice was helpful, but about 63% said they had not told anyone they were using AI this way.

• Lead author Ryan McBain of RAND said the findings point to a need for stronger regulation because chatbot mental health use is largely self-regulated under current federal law.

Why it Matters for Care

• Clinicians should consider asking adolescents and young adults directly whether they use AI chatbots for emotional support, coping advice or crisis-related questions.

• That matters because chatbot use may be filling gaps created by workforce shortages, access barriers and long waits for licensed mental health care.

• It also raises bedside concerns: overly validating responses, reinforcement of delusions, parasocial attachment and unsafe handling of suicidality or self-harm.

Between the Lines

• The appeal is not hard to parse: AI is always available, familiar, private-feeling and stigma-light for teens already using it for school or daily life.

• Some evidence suggests short-term benefit from chatbots designed for cognitive behavioral therapy, anxiety or Depression symptom support, but general-purpose bots are not the same as regulated treatment tools.

• The policy tension is that families and patients may experience chatbots as a lifeline even as experts warn the systems are optimized for engagement and affirmation, not clinical judgment.

What to Watch

• Statehouses are moving first: California and New York have passed laws requiring safeguards around suicide and self-harm, while Illinois has barred AI from being used as therapy.

• Courts may shape the field sooner than Congress, with ongoing lawsuits against OpenAI and prior litigation involving Character.AI over alleged harms to teens.

• Watch whether federal agencies or lawmakers move toward national auditing, disclosure or youth-safety standards for mental health-related chatbot use.

Source: NBC News Health