From digital companionship to real-world withdrawal: the deep impact of anthropomorphic AI on college students' social interaction and educational responses
Year: 2026
Type: article
Abstract: The rapid advancement of Artificial Intelligence (AI) is reshaping mental health services in higher education, and the widespread adoption of anthropomorphic AI is noteworthy. This article employs a literature-based theoretical reasoning approach, integrating parasocial interaction theory, self-determination theory, and uses and gratifications theory to construct a Triggering–Satisfying–Differentiating framework, systematically examining the mechanisms through which anthropomorphic AI affects college students' social psychology. The study finds that AI effectively alleviates loneliness and social anxiety in the short term by enhancing social presence, providing unconditional acceptance, and offering immediate accessibility. However, its "frictionless socializing" experience simultaneously carries three latent risks: substitute gratification weakens motivation for real-world social interaction; idealized interaction raises interpersonal expectation thresholds; and emotional dependence reinforces social avoidance behavior. For students with social anxiety tendencies, the digital haven may evolve from relief to chronic avoidance, hindering essential social skills and resilience. Accordingly, this paper proposes that universities establish "human–AI boundary" awareness, adopt a "reality-first" principle for behavioral guidance, develop connection-promoting AI tools, and build a multi-tiered digital literacy education system. The article aims to provide theoretical and practical pathways for interventions in college students' social development.
Author: Zhenyang Wu
Institution: Guangdong Polytechnic Normal University
Language: English
Subfield: Social Psychology
Field: Psychology
Domain: Social Sciences
SDG: Quality education
Open Access status: hybrid
OpenAlex ID: W7225709695
