Nearly 1 in 5 adolescents and young adults have used AI chatbots for advice or help when they felt upset, nervous or anxious, according to a new study published June 1 in JAMA Pediatrics.

Researchers also found that many of these young people use chatbots repeatedly and consider the advice helpful, suggesting that AI is already becoming part of the mental health information ecosystem for millions of young people. Those insights are raising important questions about how these tools should be used and where their limitations lie.

To help us untangle these questions, I spoke with CNN wellness expert Dr. Leana Wen, an emergency physician and clinical associate professor at George Washington University. She previously served as Baltimore’s health commissioner and is the mother of two young children.

CNN: What did this new study discover about why so many young people are turning to AI chatbots for support?

Dr. Leana Wen: This study surveyed 1,009 adolescents and young adults ages 12 to 21 across the United States and weighted the findings to represent nearly 43 million young people nationwide. Researchers asked participants whether they had used AI chatbots for advice or help when feeling sad, angry, nervous or stressed.

The team found that 19% reported doing so, which translates to more than 8 million young people nationally. Among those participants who used chatbots for emotional support, many individuals reported doing so repeatedly, with more than 4 in 10 using chatbots at least monthly. Interestingly, young people who had recently discussed their mental health with a physician were more likely to report using AI chatbots, which suggests that these tools are often being employed alongside traditional sources of support.

As to why young people are turning to chatbots, I think the answer is self-evident. These tools are available at any hour of the day, they respond instantly, they do not appear judgmental and they can feel private. For a teenager who is embarrassed to discuss a problem with a parent, teacher or counselor, typing a question into a chatbot may feel easier than starting a conversation with another person.

CNN: The study found that most users considered the advice helpful. Does that mean these chatbots are helping?

Wen: Not necessarily. This is one of the most important distinctions in the study.

The researchers found that more than 91% of users viewed the advice as somewhat or very helpful. That tells us that young people generally liked the responses they received. It does not tell us whether the responses improved their mental health, reduced symptoms of depression or anxiety, or led to better long-term outcomes.

This is a key distinction because AI chatbots are often designed to be engaging and agreeable, even flattering and sycophantic. They can make users feel heard and validated, and to want to come back for more “conversation.” Those qualities may create a positive user experience, but they are not the same as evidence-based mental healthcare.

Consider how this finding applies to other aspects of healthcare. Patients want a warm and friendly interaction with their provider, but it really matters whether the advice is accurate and actually improves someone’s health outcomes. You wouldn’t go to just anyone for your medical care, and you shouldn’t just go to a general-purpose AI tool for your mental healthcare.

CNN: What are the biggest concerns about relying on AI chatbots for mental health advice?

Wen: The first concern is accuracy. These systems can hallucinate and provide incorrect information. They can misunderstand context and offer advice that sounds convincing but is wrong and even dangerous. They are not substitutes for clinical judgment and, importantly, cannot recognize when someone may need urgent intervention.

Second, chatbots may reinforce what users want to hear rather than what they need to hear. If a young person expresses a distorted belief, unhealthy coping strategy or potentially harmful idea, the chatbot may validate parts of that thinking rather than appropriately challenging it.

Third, some adolescents may delay seeking professional help because they feel that the chatbot is providing enough support. For mild stress or everyday frustrations, choosing that type of help may not matter. For serious depression, anxiety, self-harm or suicidal thoughts, a delay in treatment matters a lot. There have been numerous cases in which parents blamed AI bots for mental health harms and suicides in their children. Teens not obtaining appropriate and timely mental healthcare can have very real consequences.

CNN: The researchers found that most young people did not tell anyone they were using chatbots for this purpose. Why is that significant?

Wen: This was one of the findings that stood out most to me. Nearly two-thirds of users said they had not told anyone they were using chatbots for emotional support.

Again, there are understandable reasons, such as privacy. The problem is that parents, clinicians and other trusted adults may have no idea how much AI is influencing how a young person thinks about emotional challenges. If an adolescent is receiving inaccurate information or becoming overly dependent on a chatbot, the adults in their life may never know.

That is why I think parents and healthcare professionals should begin asking about AI use in the same way they ask about social media use. It is becoming an increasingly important part of how young people gather information and seek support.

CNN: Are there situations where AI chatbots can play a positive role in emotional support or mental health?

Wen: Possibly, yes. For instance, AI chatbots may help some people practice difficult conversations, learn basic coping strategies or identify mental health resources. They may also lower barriers for people who are hesitant to seek support. For example, a teenager experiencing anxiety might use a chatbot to learn about relaxation techniques or to develop questions to ask a therapist.

Some AI tools are being built specifically to help patients who are already under the care of a mental health professional. If validated in the future, they may become better alternatives to the generic AI chatbots that we currently have. Again, though, the danger is when AI becomes a substitute for human relationships and professional care.

CNN: What can parents do to support their child’s emotional well-being when AI is becoming part of everyday life?

Wen: I recommend approaching AI much as many families have learned to approach social media. Understand what tools your child is using. Ask open-ended questions. Stay curious rather than judgmental.

Parents can help children understand that AI has strengths and limitations. They can query AI together and see how chatbots can provide helpful suggestions but also emphasize how important it is to verify the information provided and to question responses. That includes typing the same question in different chatbots. And they can certainly emphasize that AI is a tool, not a replacement for a therapist or trusted family member.

I also would encourage parents to model healthy help-seeking behavior. Children should know that asking for support is a sign of strength, not weakness. They should understand that emotional challenges are a normal part of life and that there are many trusted — human — sources of help available, including teachers, counselors, pediatricians, and, of course, parents and other caregivers.

CNN: What is your advice for adolescents and young adults who are struggling with stress, anxiety, sadness or other mental health concerns?

Wen: If you are struggling, please know that you do not have to handle it alone. Speak to your parents or other trusted adults. If symptoms are persistent, interfere with daily life or involve thoughts of self-harm, professional help is especially important. That means contacting your primary care physician or licensed mental health professional. If you don’t know where to turn, call 988, the free and confidential 24/7 hotline where you will be quickly connected to a mental health expert who can help.


Contributor: Katia Hetter – CNN Health