Frequency of use and attitudes toward AI-based health counseling among patients attending the dermatology outpatient clinic: a cross-sectional study.

PURPOSE: To quantify the prevalence of pre-visit AI consultation among dermatology outpatients and to assess associated attitudes, anxiety changes, and triage impacts, given that real-world data on generative artificial intelligence (AI) use in dermatology remain scarce. MATERIALS AND METHODS: A cross-sectional survey was conducted at a university dermatology clinic. Data on demographics, AI use characteristics, and attitudes (7-item Likert scale) were collected. Multivariable logistic regression identified predictors of AI use. RESULTS: Among 349 patients (median age 25 years), 61.1% (n = 212/347; 95% CI: 55.9-66.1%) had consulted AI before their visit, predominantly using ChatGPT (77.4%). Younger age (aOR = 0.93) and frequent internet health searching (aOR = 3.43) independently predicted AI use. AI consultation was not associated with a detectable change in self-reported anxiety (Wilcoxon p = 0.836; rank-biserial r = 0.03). For 73.6% of users, AI confirmed an existing decision to seek care, while 17.5% visited because AI indicated severity. Patients strongly rejected AI as a physician substitute (mean 1.66/5) but partially endorsed it for asking informed questions (mean 2.89/5). Patient-recalled symptom-diagnosis plausibility, evaluable in a 33.5% recall subset, was 83.8% (57/68). CONCLUSIONS: Generative AI use is highly prevalent among younger dermatology outpatients. Within the limits of a cross-sectional design, it appeared to function primarily as a confirmatory pre-visit resource and was not associated with increased health anxiety. Patients view AI as an empowerment resource rather than a diagnostic authority.