ABSTRACT
This chapter argues that Generative AI (GenAI) systems blur the boundary between tool and social actor, in some cases prompting users to engage with them as though they were emotionally responsive, socially meaningful agents. Drawing on frameworks of multidimensional agency, evolving social norms, and simulated reciprocity, the chapter explains how users interpret GenAI as capable of adaptive, relational interaction. It introduces the concept of AI Social Interaction (AISI) to describe these hybrid engagements and explores their implications through empirical and theoretical lenses. Finally, it identifies ethical concerns that arise when users form emotional attachments to AI, particularly in light of perceived reciprocity and authenticity. The chapter calls for ethical oversight, transparent system design, and new theoretical tools to address the relational, emotional, and social complexity of human-AI interaction.
