Asian Journal of Business Ethics Philosophy

ChatGPT as a digital companion: Ethical reflections on constructive use, social connection, and mental health in Singapore

Wong, Ada Hiu Kan, Lee, Carmen Kar Hang
Springer Nature · 2026-09-15 · DOI 10.1007/s13520-026-00271-1

Abstract

As generative artificial intelligence (AI) tools such as ChatGPT become increasingly embedded in everyday life, important ethical questions arise about their implications for human well-being, autonomy, and social connection. Drawing on survey data from Singapore ( N = 1,143), this study applies association rule mining to explore how ChatGPT use experiences co-occur with self-reported depression severity. The findings show that respondents who perceived ChatGPT as reducing stress, enhancing work efficiency, or increasing motivation to learn, while not reporting increased loneliness or disconnection, were more likely to report none-to-mild depressive symptoms. The obtained rules indicate that lower depression severity co-occurred with preserved confidence in handling daily life issues, critical thinking, and creativity. Framed by the Work-Home Resource model, we interpret these findings as evidence that ChatGPT may function as a contextual digital resource when its use reinforces users’ self-efficacy, cognitive agency, and social connectedness. At the same time, the findings highlight ethical concerns around over-reliance, one-sided AI companionship, transparency, human dignity, and responsible deployment. The study contributes to the emerging discussions on AI ethics and digital well-being by showing how constructive AI use, social connectedness, and cognitive confidence jointly shape the well-being implications of conversational AI. Practical implications are discussed for organisations, technology designers, policymakers, and digital literacy initiatives, particularly in the Asian context.

Read at publisher →