ABSTRACT In an increasingly digitalized era characterized by heightened emotional and social needs, artificial companions (ACs) have emerged as alternative social actors that provide affective interaction and companionship. Although prior research has examined system proactivity in task‐oriented contexts, limited attention has been paid to companion‐oriented systems that emphasize ongoing social interaction. Addressing this gap, the present study investigates how AC proactivity shapes users' willingness to use through perceived social presence in affective companionship contexts. Using a mixed‐methods experimental approach, we find that AC proactivity significantly enhances willingness to use, and this effect is fully mediated by perceived social presence. Furthermore, loneliness moderates the relationship between AC proactivity and social presence: users with lower loneliness experience stronger social presence in response to proactive interaction, whereas highly lonely users may exhibit resistance or heightened vigilance. By clarifying the psychological mechanisms and boundary conditions of proactive conversation initiation, this study advances understanding of proactivity as a social cue in human–computer interaction and offers implications for the context‐sensitive design of companion‐oriented AI systems.
Li et al. (Thu,) studied this question.