Greater mental health use and concern about Generative AI among college students with anxiety-depression

Objective: To examine whether generative artificial intelligence (GenAI) use and AI-related sentiment differ across anxiety–depression symptom groups among college students. Participants: 890 U.S. college students who completed the elective AI module of the 2024–2025 Healthy Minds Study. Methods: Participants were grouped into Anxiety–Depression, Anxiety Only, Depression Only, and Neither using GAD-7 and PHQ-9 scores (≥10). ANCOVAs with Bonferroni-adjusted comparisons examined AI use (overall, mental health, companionship) and AI-related sentiment, adjusting for sociodemographic covariates. Results: Overall AI use did not differ across groups. Mental health-related and companionship AI use differed across groups; the Anxiety–Depression group reported greater mental health-related use than the Neither group (p < .01). The Anxiety–Depression group also reported greater concern about AI than the Neither group (p < .01). Conclusions: Differences across anxiety and depression symptom groups were evident for specific forms of GenAI use and AI-related concern, but not for overall use.