The imaginary nature of human-AI relationships: a perspective integrating Buddhist psychology and the psychology of anthropomorphism

Abstract

The rapid growth of AI companionship has raised urgent questions about the psychological mechanisms underlying human-AI relationships. This paper investigates why and how users construct and become entrenched in relationships with AI. Drawing on the Consciousness-Only school of Buddhist psychology, this paper applies the three-natures framework and eight-consciousnesses model. According to this framework, our minds tend to superimpose constructed meanings and characteristics on dependently arising phenomena. Anthropomorphism of AI systems is a manifestation of the imaginary nature. Studies on anthropomorphism indicate that users project mentality onto AI systems due to emotional and social needs. However, the Consciousness-Only school reveals a deeper root: a craving for a stable relational object through which the self is secured. Current AI systems may satisfy this deeper craving because they are perceived as always responsive, unconditionally accepting, and free from relational conflicts. These perceived qualities reside in the users’ perceiving minds, rather than in the AI systems. Because this craving operates beneath conscious awareness, informational correction (“AI is based on Large Language Models”) is not sufficient to dissolve the projection. Moreover, without relational frictions, self-grasping and the imaginary nature cannot be recognized and relinquished. The framework therefore suggests that increasing reliance on AI systems may contribute to social isolation. Neither AI usage nor anthropomorphism is inherently problematic, but ignorance of this projection and reliance on it are. Instead of avoidance, this paper calls for cultivating awareness of our own projections and the imaginary nature, which could interrupt the self-reinforcing loop and support healthy human-AI relationships.