For many young people, asking an AI chatbot for emotional advice is no longer unusual. A nationally representative US study published in JAMA Pediatrics in 2026 found that 19.2% of adolescents and young adults aged 12 to 21—about 8.2 million people—had used AI chatbots for mental health advice in 2025. That was up from 13.1% in a similar national survey published in 2025. Among the 2025 users, 42.8% said they turned to chatbots at least monthly, and 91.7% described the advice as somewhat or very helpful. (jamanetwork.com)
The appeal is easy to understand. Earlier JAMA researchers suggested that young users may like AI advice because it feels low-cost, immediate, and private. But the newer study also found a quieter, more troubling detail: 63.3% of users told no one that they were using AI for mental health advice. Use was more common among females, among 18- to 21-year-olds, and among those who had recently spoken with a physician about their mental health. In other words, AI is becoming, for some young people, a hidden source of emotional support. (jamanetwork.com)
Still, “helpful” does not always mean “safe.” A 2025 systematic review and meta-analysis of 31 randomized trials found small-to-moderate overall benefits from AI chatbots for reducing mental distress in adolescents and young adults, but it also said the evidence for generative systems remained inconclusive and called for better safety rules and more long-term research. The World Health Organization echoed this concern in March 2026, noting that generative AI tools are increasingly being used for emotional support by young people even though they were not designed or tested for mental health care. (doi.org)
Recent survey data suggest that this habit may spread even earlier in life. In Common Sense Media’s 2026 census, 37% of AI-using children aged 9 to 17 said they had used AI to discuss their feelings or personal problems, and 12% of all children surveyed said they would ask an AI chatbot before an adult for information about their health or body. The same report found that 17% of young chatbot users had seen something they felt was not appropriate for their age. So the real question is no longer whether young people will talk to AI. It is whether adults, schools, and health systems can make those conversations safer. (commonsensemedia.org)










