
The use of general-purpose AI chatbots (i.e. ChatGPT, Claude, Gemini, etc.) continues to expand, with surveys suggesting that about half of Americans use these tools. While AI chatbots were developed to provide human-like, conversational assistance with tasks such as finding information and managing work, many people, including adolescents and young adults, turn to AI chatbots for emotional support and mental health assistance. These AI chatbots were not developed, designed, or intended for these mental health uses, and national organizations have highlighted concerns, including suicide risk, when AI chatbots are used in this unintended manner. Here are some important points on this topic:
Risks of AI chatbots for mental health concerns
- AI chatbots cannot substitute for a trained mental health professional. National organizations caution that AI chatbots are not regulated or tested for mental health use. AI chatbots are designed to maximize user engagement – in contrast to human mental health professionals who have ethical and legal obligations to prioritize a patient’s well-being and safety.
- Use of AI chatbots can evolve into unhealthy dependence and reinforce mental health symptoms related to depression, obsessive compulsive disorder, mania, psychosis, eating disorders, and suicide.
- AI chatbots do not identify or respond reliably to mental health crises. A 2023 UW study found that AI chatbots did not consistently recommend human help until risk was very high, and few provided suicide hotline information.
Our recommendations for AI chatbot use
- Use AI chatbots for their intended purposes: Accessing information, organizing work tasks, creating journaling prompts – NOT mental health support, diagnosis or crisis care.
- Treat it as a tool, not a therapist: Turn to trusted support people or a professional for emotional support and mental health needs.
- Protect your privacy: Avoid sharing identifying details, passwords, or anything you would not want stored or reviewed. Conversations with a therapist are legally confidential; conversations with a bot are not.
- Verify advice: Always verify urgent advice with a clinician, counselor, or crisis line before acting on it. · Know how to access crisis resources: The 988 Suicide & Crisis Lifeline (phone and text) and Crisis Text Line (text 741-741) are available for 24/7 support with a trained crisis counselor.
References
American Psychological Association. (2025). Health advisory: Use of generative AI chatbots and wellness applications for mental health. https://www.apa.org/topics/artificial-intelligence-machine-learning/health-advisory-chatbots-wellness-apps
Common Sense Media. (2025). Common Sense Media AI Risk Assessment: AI Chatbots for Mental Health Support. https://institute.commonsensemedia.org/sites/default/files/ra-import/csm-ai-risk-assessment-mental-health-11142025.pdf
Heston, T.F. (2023). Safety of Large Language Models in Addressing Depression. Cureus, 15. https://doi.org/10.7759/cureus.50729
McBain, R. K., et al. (2026). AI Chatbot Use and Disclosure for Mental Health Among US Adolescents and Young Adults. JAMA Pediatrics, 180(8), 884–890. https://doi.org/10.1001/jamapediatrics.2026.2015
Pew Research Center. (2026). Americans and AI 2026: Chatbots, Smart Devices and Views on Impact. https://www.pewresearch.org/wp-content/uploads/sites/20/2026/06/PI_2026.06.17_Americans-and-AI_REPORT.pdf
The JED Foundation. (n.d.). Why Your Teen Shouldn’t Be Using AI Companions – and What to Do If They Are. https://jedfoundation.org/why-your-teen-shouldnt-be-using-ai-companions-and-what-to-do-if-they-are/







