Main Article Content

Abstract

Background: Large-language-model AI companions invite reciprocal-seeming parasocial relationships, yet no integrated model explains why relational strain translates into machine attachment in a Global-South setting.


Objective: To test whether social-interaction burnout and loneliness predict attachment to AI companions, whether parasocial interaction mediates the burnout-attachment path, and whether collectivist judgment apprehension moderates the loneliness-attachment link.


Methods: A cross-sectional survey of 1,200 young adults recruited through a public organization in Palembang measured burnout, loneliness, judgment apprehension, AI parasocial interaction, and emotional attachment using validated scales.


Results: The model explained 48% of attachment variance. Parasocial interaction, loneliness, and burnout were the strongest predictors. Parasocial interaction partially mediated the burnout-attachment path, and judgment apprehension strengthened the loneliness-attachment association.


Conclusion: The findings support an 'Algorithmic Sanctuary' account of AI companionship and inform digital-wellbeing policy in collectivist societies.

Keywords

AI companions Emotional labor Indonesia Loneliness Parasocial interaction

Article Details

How to Cite
Ni Made Nova Indriani, Immanuel Simbolon, & Sophia Lucille Rodriguez. (2026). The Algorithmic Sanctuary: Social-Interaction Burnout, Loneliness, and Emotional Attachment to AI Companions Among Young Adults in Indonesia. Open Access Indonesia Journal of Social Sciences, 9(3), 138-145. https://doi.org/10.37275/oaijss.v9i3.327