A thematic analysis of users’ experience of an AI-enabled mental health intervention for generalised anxiety symptoms

Anxiety disorders are highly prevalent, yet access to effective psychological treatments remains limited. Artificial intelligence (AI)-enabled mental health applications offer a scalable and accessible means to bridge treatment gaps. This study qualitatively explored the experiences of adults with anxiety symptoms using PATH, an AI-enabled mental health intervention. Participants (N = 100) with moderate-to-severe anxiety were recruited via Prolific and engaged with PATH before providing feedback. A total of 67 participants completed the qualitative questions at the 2-week timepoint, and 51 participants completed a further questionnaire at the 8-week timepoint. Participants had a mean age of 39.84 (SD = 11.19). Most were born in the UK (89.6%), and reported their ethnicity as White (92.5%), Asian (3.0%), mixed (3.0%) and Black (1.5%). Many worked full-time (53.7%) or part-time (20.9%), and a minority (14.9%) were students. Data were analysed using Qualitative Descriptive Analysis. Four overarching themes emerged: (1) Key Reasons for Use; (2) Usability and Functionality; (3) Preferred In-App Interventions; and (4) Facilitators and Barriers to Sustained Engagement. Findings underscore the value of empathetic, personalised and readily available AI-enabled interventions for anxiety management, while also highlighting design challenges requiring refinement.