ABSTRACT Benefiting from advances in artificial intelligence technology, social chatbots account for changing the way human–robots socialize, especially improving their psychological state. This paper aims to not only establish the relationship between loneliness and attachment to social chatbots but also explore the affecting mechanism. The methods of machine learning (Study 1) and experiments (Studies 2 and 3) are used to explore and validate the affecting mechanism. The results demonstrate that loneliness can positively promote users' attachment to social chatbots, and this process is mediated by self‐disclosure. Furthermore, it is worth noting that perceived empathy moderates not only the effect of loneliness on self‐disclosure but also the effect of loneliness on attachment to social chatbots. This research contributes to understanding human–social chatbot dynamics, providing valuable insights for marketers and firms involved in designing coding programs for enhanced user experiences during interactions with social chatbots.
Yao et al. (2025) studied this question.