Abstract
Importance: Depression and anxiety are highly prevalent and disabling, yet most affected individuals receive inadequate or no care. Purpose-built generative AI platforms may offer scalable first-line mental health support, though large-trial evidence remains limited.Objective: To evaluate whether Flourish, an AI-powered mental health platform, reduces depression and anxiety symptoms over six weeks.Design, Setting, and Participants: This preregistered randomized clinical trial included 1,137 college students randomized to Flourish or care-as-usual. Participants were recruited without symptom-based screening and assessed at baseline and weeks 2, 4, and 6.Interventions: Flourish, an AI-powered platform delivering CBT-, DBT-, ACT-, positive psychology–, and motivational interviewing–informed support through an AI well-being coach, Sunnie, vs care-as-usual.Main Outcomes and Measures: Depression and anxiety symptoms, assessed using the PHQ-2 and GAD-2.Results: Flourish produced more favorable symptom trajectories than control for depression (condition × time: β = −0.043; p = .006) and anxiety (β = −0.058; p = .001). Consistent with Flourish’s intended role as a first-line support tool, favorable patterns were observed across the baseline symptom spectrum, with less symptom worsening among participants with low baseline symptoms and greater symptom reduction among those with elevated symptoms. Among participants with elevated baseline symptom severity, Flourish produced higher rates of meaningful change for depressive symptoms (52.3% vs 31.7%; NNT = 4.9) and anxiety symptoms (50.6% vs 38.1%; NNT = 8.0).Conclusions and Relevance: To our knowledge, this is the largest randomized clinical trial of a conversational AI mental health platform and among the first to demonstrate effectiveness in a population-based college sample recruited without symptom-based screening. These findings support purpose-built generative AI platforms as scalable first-line mental health tools.