The mental health crisis among young people is on the rise. Unfortunately, limited access to professional help still remains one of the largest roadblocks to effectively dealing with mental health issues. Soon after AI chatbots entered the market, many people began turning to tools like ChatGPT or Google Gemini for more than just information. They started leaning on these generative AI models for emotional reassurance and everyday psychological support.
A recent study set out to better understand how often and in what ways adolescents and young adults in the US were turning to AI chatbots for mental health advice. The online survey estimated that roughly 1 in 5 adolescents and young adults in the US have used an AI chatbot for mental health advice as of 2025, and most users found the advice helpful.
Despite finding support through these tools, a majority of users, 63.3% of them, chose to keep seeking help from a chatbot hidden from others.
The findings are published in JAMA Pediatrics.
Parallel rise of crisis and chatbots
According to recent CDC data, about 1 in 3 U.S. high school students reported poor mental health most or all of the time over the past 30 days. In a 2023 survey, 20.4% said they had seriously considered suicide, while nearly 1 in 10 reported at least one attempt at taking their life. Rates of depression and anxiety disorders among adolescents also remain high. Together, these numbers point to a worrying picture of youth mental health in the US.
AI chatbots are rapidly gaining popularity at the same time the youth mental health crisis continues in the US, prompting researchers to question how much young people are using them for support. While previous studies have established the use of AI for seeking mental health advice, there is very little data about the usage patterns among young adults and adolescents.
For this study, researchers invited 1,727 young people in the United States, aged 12 to 21, to take an online survey. A total of 1,009 young people completed the survey.
To ensure the results accurately reflected the entire U.S. population in that age group, the researchers applied a statistical technique called weighting, which allowed the 1,009 responses to represent over 42 million youth nationwide.
The participants were asked four major questions:
- a) If they had ever used an AI chatbot for advice when feeling sad, angry, nervous, or stressed
- b) If they said yes, how often they used the chatbot
- c) Was the advice helpful?
- d) Did they tell a parent or friend, or anyone, about using AI for this purpose?
Nearly 20% of young people answered yes to using AI for mental health support, a significant jump from a previous survey conducted a year ago, in which only 13% replied affirmatively to the use of AI for this purpose. More than 40% of users turned to AI chatbots for advice at least once a month and 5.8% reported using them daily or almost daily.
Yet a majority of those who used the chatbot for help at any point kept their use a secret. 91.7% said they found the experience helpful. Researchers, however, caution that this sense of helpfulness may stem from chatbots’ tendency to be overly agreeable or flattering, rather than from the quality or accuracy of the advice itself.
Among the demographics, females and older teens were more likely to use AI chatbots for support. Usage was also higher among those who had spoken to a physician about their mental health within the previous six months.
The researchers note that AI chatbots have already become an integral part of many young people’s mental health information ecosystem. As their use continues to grow, parents and clinicians need to openly discuss chatbot use to help young people stay safe, set realistic expectations, and connect with mental health professionals when needed.