According to APA’s 2026 Chatbots and Mental Health Survey, 77% of psychologists say their patients report using AI. More than a third of psychologists say their patients are using AI as an additional mental health provider.
Asking patients about their AI use and creating room to explore what they are finding is an opportunity to help patients stay safe and find accurate, research-backed information.
This general evidence-based guidance on safe use is a helpful foundation for building your knowledge and framing those conversations.
Here are practical suggestions for approaching these AI discussions in a curious, patient-centered way—drawing on recommendations from APA’s health advisory on use of generative AI chatbots and wellness applications for mental health, recent media, and other scholarly commentary aligned with APA’s advisory.
- Proactively ask patients about their AI use. Normalize that many people are using AI for support and inquire if the patient has tried any AI tools.
- Explore how the patient has found the tool to be helpful or unhelpful.
- Ask whether you can share some information about AI. If yes, share information and/or concerns that may be relevant, such as noting that AI can give incorrect information and is not suitable support in a crisis or emergency.
- Invite ongoing conversations about the patient’s AI use, including possibly bringing outputs to session to discuss. Consider adding a question about AI use on intake forms.
- Help patients establish clear boundaries for how and when they use these tools, and warning signs of unhealthy use to watch out for, such as feeling the need to consult AI before making any decision, spending less time socializing in real life in favor of engaging with AI, or feeling like AI knows you better than anyone else.
- Know and discuss with patients what information is and isn’t safe to share with generative AI chatbots and apps. These platforms do not offer the privacy protections of a clinical relationship, and personal information entered into an AI chatbot may be stored, reviewed, or otherwise used by the platform.
- Talk with patients about algorithmic bias and limitations. Explain how these chatbots may propose inappropriate interventions for certain groups or provide inaccurate diagnoses or other harmful advice.
- Emphasize that AI is not a crisis resource.
