Related Experiment Video
Updated: Jun 26, 2026
07:14
Virtual Agent for Real-Time Motivational Interviewing by Integrating Adaptive Nonverbal Behavior and Language Models
Published on: December 23, 2025
Supporting Student Mental Health With the Safespace Generative AI Chatbot: Mixed Methods Feasibility Study.
Matteo Pinna1,2, Sergio Galletta1,3, Elliott Ash1
1Center for Law and Economics, Department of Humanities, Social and Political Sciences, ETH Zürich, Haldeneggsteig 4, Zürich, 8092, Switzerland, +41 44 6324586.
JMIR Formative Research
|June 24, 2026
Summary
Generative artificial intelligence (GenAI) chatbots show promise for scalable mental health support. The Safespace GenAI chatbot was feasible and used frequently by students, particularly for emotional disclosure and between-session reflection.
Area of Science:
- Mental Health Technology
- Artificial Intelligence in Healthcare
- Digital Mental Health Interventions
Background:
- Generative artificial intelligence (GenAI) chatbots offer potential for large-scale, personalized mental health support.
- The Safespace chatbot is an AI-driven smartphone application utilizing a large language model for mental health assistance.
Purpose of the Study:
- To evaluate the feasibility and usage patterns of the Safespace GenAI chatbot among university students.
- To explore user attitudes toward GenAI chatbots and analyze qualitative experiences.
Main Methods:
- A mixed-methods approach was employed, including surveys, qualitative content analysis, and descriptive assessment of preintervention depressive symptoms.
- The study involved 42 university students, with 20 actively using the chatbot over 2-4 weeks, resulting in 286 interactions.
Main Results:
- Most participants found the chatbot helpful (64%) and trusted its privacy (93%).
- Peak usage occurred during early morning and late night hours, potentially when other support is unavailable.
- Participants used the chatbot for reflection as a blended-care tool, though technical barriers and design needs were identified. Elevated depression scores correlated with higher emotional disclosure during sessions.
Conclusions:
- The Safespace GenAI chatbot demonstrates feasibility and engagement potential for mental health support.
- Findings highlight the need for further research to optimize GenAI interventions, addressing technical aspects and user engagement for sustained use.