Would You Trust Your GenAI Chatbot as Your School Therapist? Perspectives and Implications From Teachers and Students

Abstract

This study examines trust in AI chatbot therapists within the context of Yang and Wibowo’s conceptual framework on user trust in generative artificial intelligence (GenAI), which addresses factors, components, and outcomes of trust. Drawing on insights from secondary school teachers ( n = 29) and students ( n = 69), the research evaluates the opportunities and challenges of integrating GenAI into mental health support services in school settings. Qualitative data analysis was conducted using deductive coding guided by Yang and Wibowo’s framework on user trust in AI, followed by inductive coding to capture themes that fell outside the predefined categories. Participants appreciated the accessibility, predictability, and nonjudgmental nature of AI, highlighting its potential to enhance accessibility to these services. However, concerns about data security, emotional detachment, and the inability of AI to address complex emotional issues limited deeper trust. The perceived ability, integrity, and benevolence of AI emerged as critical, with participants achieving cognitive trust in AI to complete basic tasks but affective and long‐term trust remaining elusive. Findings highlight the societal and educational implications of AI adoption for mental health, emphasizing the need for enhanced emotional intelligence of AI tools, transparent data governance, and the positioning of AI as a complementary tool to human therapy. This research offers practical insights for integrating AI responsibly in mental health services while addressing broader societal challenges.

Ähnliche Arbeiten