Authors: Chumki Bose
Abstract: Background Generative artificial intelligence (GenAI) has rapidly moved from an information and productivity technology into an interpersonal technology through conversational systems capable of sustained, personalized and emotionally responsive interaction. Young adults increasingly use conversational artificial intelligence (CAI) for reflection, advice, emotional disclosure, problem-solving and companionship. Although emerging international research suggests that conversational AI may reduce aspects of psychological distress and loneliness for some users, concerns have also been raised regarding emotional dependence, displacement of human relationships, privacy, inappropriate psychological advice and reinforcement of maladaptive beliefs. Evidence from India remains limited despite the country's rapid expansion of digital mental-health services and large young-adult population. Objective The proposed study aims to examine the psychological factors associated with conversational AI use for emotional support among Indian young adults. Specifically, it will investigate whether loneliness and psychological distress predict greater use of conversational AI for emotional purposes, whether perceived social support moderates this relationship, and whether patterns of emotional AI use are associated with conversational AI dependence. Method A cross-sectional mixed-methods study is proposed involving Indian adults aged 18–29 years who report using a conversational AI system at least once during the preceding three months. A target sample of 600–800 participants will be recruited using stratified online sampling across major Indian regions. Measures will include loneliness, psychological distress, perceived social support, wellbeing, conversational AI use characteristics and AI dependence. A validated conversational-AI-dependence measure will be adapted and psychometrically evaluated for the Indian context. Quantitative data will be examined using hierarchical regression and structural equation modelling, while qualitative responses concerning reasons for AI use, emotional disclosure, perceived benefits and perceived risks will undergo reflexive thematic analysis. Expected contribution The study is designed to distinguish between AI use as a compensatory form of emotional support and AI use as a potentially maladaptive substitute for human social connection. The findings could inform culturally responsive digital mental-health policy, psychological assessment, responsible AI design and integration of conversational technologies with India's emerging digital mental-health infrastructure.
