A client walks into a session and says, “I asked AI about my symptoms, and it says I might have ADHD.” Another shares a chatbot’s explanation of a relationship conflict. Someone else arrives with a coping strategy generated online or a list of reasons they believe they have a particular mental health condition.
For behavioral health professionals, these conversations are becoming increasingly common.
According to the American Psychological Association’s 2026 survey, 77% of psychologists reported that patients had discussed using AI, including for emotional support, self-diagnosis, and other mental health concerns. The findings highlight an important question for clinicians: how can professionals respond when AI becomes part of a client’s understanding of their own mental health?
Start With Curiosity, Not Criticism
When a client brings AI-generated information into a session, the first response matters.
Dismissing the information outright may leave the client feeling embarrassed or misunderstood. Agreeing with it without examining its accuracy can reinforce an incorrect conclusion.
Instead, clinicians can approach the conversation with curiosity.
Questions such as “What stood out to you about that response?” or “How did reading that make you feel?” can help reveal what the client is seeking. They may be looking for reassurance, trying to understand confusing symptoms, or searching for an explanation for a difficult experience.
The AI-generated answer may not be clinically accurate, but the client’s reaction to it can provide useful information.
The goal is not to debate technology. It is to understand the client’s concerns and help them evaluate the information thoughtfully.
AI Can Sound Certain Without Being Correct
One challenge with generative AI is that its responses can sound confident and convincing, even when the information is inaccurate, incomplete, or poorly suited to the individual.
A chatbot may suggest a diagnosis based on a short description of symptoms without considering medical history, other possible explanations, duration, severity, or effects on daily functioning.
Similarly, it might interpret a partner’s behavior as manipulative or recommend a coping strategy without understanding the broader circumstances.
The American Psychological Association cautions that general-purpose chatbots are not a reliable way to diagnose mental health conditions. Their responses may also reinforce a user’s existing assumptions instead of challenging them.
Clinicians can help clients understand that AI-generated information is a starting point for questions, not a substitute for professional assessment.
Ask What the Client Hopes to Gain
AI use often reflects an underlying need.
A client seeking a diagnosis may want validation after years of feeling misunderstood. Someone asking a chatbot for relationship advice may be unsure how to communicate their needs. A person looking for coping strategies may feel overwhelmed between appointments.
Understanding the motivation behind the search helps clinicians respond to the actual concern rather than focusing only on whether the AI’s answer was correct.
Professionals might ask:
- What were you hoping AI would help you understand?
- Did the response reassure you, worry you, or leave you with more questions?
- Does the information fit your experience?
- Was there anything in the answer that felt inaccurate or uncomfortable?
- How would you like to work through this together?
These questions keep the therapeutic relationship at the center of the conversation.
Evaluate Advice Against Clinical Knowledge
Not every AI-generated suggestion is harmful. A chatbot might help a client organize thoughts before an appointment, brainstorm questions for a clinician, or remember a coping exercise already discussed in treatment.
The important distinction is between using AI to support an established care plan and relying on it to make clinical decisions.
If a client shares a new coping technique, the clinician can discuss whether it is appropriate for the client’s needs and circumstances. If the client arrives with a diagnosis, the professional can explain how a proper assessment works and consider relevant symptoms through established clinical practice.
AI-generated relationship advice should also be examined in context. A chatbot typically receives only the information a person provides, which may reflect one perspective or omit important details.
Clinicians can help clients consider alternative explanations, identify missing information, and evaluate possible consequences before acting on the advice.
Address Privacy and Safety
Clients may not fully understand what happens to information they share with AI platforms.
They might enter details about trauma, substance use, family conflict, medication, or other highly personal experiences. Depending on the service and its settings, that information may be stored or handled in ways the client does not expect.
Professionals should encourage clients to review the privacy practices of the tools they use and avoid entering identifying or sensitive health information into services without understanding how it may be handled.
Safety also deserves direct attention. If AI-generated content encourages self-harm, reinforces a dangerous belief, or recommends stopping treatment or medication without medical guidance, clinicians should address those concerns promptly and follow appropriate clinical and crisis protocols.
Keep the Human Relationship Central
AI may offer immediate responses, but it does not replace the professional judgment, accountability, and individualized understanding involved in behavioral health care.
A client may appreciate a chatbot because it is available at any hour and feels nonjudgmental. Those experiences can be meaningful, particularly for someone who struggles to ask for help.
At the same time, clients may become overly reliant on AI for reassurance or use it to reinforce beliefs that contribute to distress. The APA’s 2026 survey identified concerns about dependency and the potential reinforcement of unhelpful or distorted thinking.
Clinicians can discuss these patterns without shaming clients. The question is whether the tool is helping the person function, reflect, and engage in treatment, or whether it is increasing distress, avoidance, or dependence.
Make AI Part of the Conversation
Behavioral health professionals do not need to have every answer about emerging technology to begin discussing it with clients.
They can ask about AI use during relevant conversations, encourage clients to verify health information, and help them distinguish general information from individualized clinical guidance. The APA also recommends that practitioners create space for patients to discuss their AI use openly.
Ultimately, the most useful response to a client who brings AI into the room may be a simple question: “Let’s look at that together. What does it mean to you, and how does it fit with what you’re experiencing?”
That approach respects the client’s effort to understand themselves while keeping care grounded in professional knowledge, thoughtful evaluation, and the therapeutic relationship.