Patients are increasingly using generative AI as a mental health resource—often between or instead of clinical encounters—making it important for clinicians to routinely ask about and integrate this use into care. AI interactions can reveal otherwise undisclosed concerns and shape patients’ thinking, but they also carry risks such as inaccuracies, bias, lack of crisis response, and privacy limitations. Clinicians do not need deep technical expertise, but should apply patient-centered communication to explore how individuals use AI, reinforce benefits, address harms, and guide safe, critical engagement as part of ongoing therapeutic work.