User Preferences for AI-Powered Mental Health Chatbots in Malaysia: A Thematic Analysis
AI-powered mental health chatbots are increasingly positioned as scalable solutions to address gaps in mental health service accessibility. However, existing research remains largely Western-centric and system-oriented, with limited insight into how users in culturally diverse contexts construct meaning and form preferences toward these technologies. This study addresses this gap by examining how Malaysian users evaluate and explain their preferences for AI-powered mental health chatbots designed for the local context. A qualitative research design was employed using thematic analysis of 470 open-ended survey responses from Malaysian users. Data analysis followed a six-phase interpretive framework, supported by AI-assisted coding and pattern recognition, with continuous researcher validation to ensure analytic rigor and contextual accuracy. The analysis identified three interrelated dimensions shaping preference formation: cultural alignment, functional accessibility, and emotional safety. These dimensions function as interpretive lenses through which users assess the contextual relevance, practical value, and psychological acceptability of AI-based mental health support. Preference formation is not uniform but contingent and negotiated. Acceptance is shaped by expectations of system reliability and informational adequacy, whereas resistance emerges from concerns regarding emotional authenticity, trust, and the perceived limits of AI empathy. The study advances a culturally grounded, user-centered perspective on AI chatbot evaluation by foregrounding meaning-making processes in non-Western contexts and offers design implications for developing culturally adaptive and emotionally responsive digital mental health systems.
Keywords: AI-powered mental health chatbots; cultural alignment; emotional safety; user meaning-making; thematic analysis; Malaysia
