Is Interaction Between Human and Artificial Intelligence-Driven Agents (Para)Social? A Scoping Review.

Abstract

Despite the growing presence of artificial intelligence (AI)-based agents in human daily life, there remains a dearth of exploration and synthesized evidence on the social aspects of human-AI interaction. Researchers have increasingly viewed AI agents as a distinct form of media persona, sparking interest in the phenomenon of parasocial interaction (PSI) with these entities. Building on this emerging area of inquiry, this scoping review aims to examine the applicability of the parasocial concept to human-AI agent interactions and to evaluate whether these interactions can be considered "parasocial." Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, the findings suggest a propensity to misconstrue PSI in terms of reciprocity and sociability within the context of enhanced interactivity in human-AI communication. This study contributes to parasocial theory building and offers critical insights for future human-AI PSI research, underscoring the importance of applying the parasocial concept within its intended conceptual boundaries, particularly in the field of human-AI and broader human-machine communication.