From Algorithms to Affection: Unpacking the Psychological Mechanism of Emotional Attachment Toward Virtual Influencers
As virtual human influencers (VHIs) increasingly inhabit social media landscapes, they are functioning not merely as marketing tools but as social entities capable of eliciting user affection. However, the psychological mechanisms underlying emotional attachment to these artificial agents remain underexplored compared to their human counterparts. This study investigates how VHI self-disclosure influences emotional attachment, positing parasocial relationships as a mediator and perceived attractiveness and humanness as boundary conditions. We analyzed online survey data from 540 young adults familiar with VHIs using a moderated mediation framework. The findings reveal a complex dynamic distinct from human interaction: VHI self-disclosure generally hindered parasocial relationship formation unless the VHI was perceived as physically attractive. This suggests attractiveness acts as a requisite gateway for users to accept artificial disclosure as relationally meaningful. Subsequently, parasocial relationships strongly predicted emotional attachment, an effect significantly amplified by perceived humanness. Consequently, self-disclosure indirectly promoted emotional attachment only when the VHI was viewed as both attractive and human-like. These results delineate the distinct roles of VHI attributes: Attractiveness facilitates the initial establishment of a social connection (parasocial relationship), whereas perceived humanness deepens this bond into a meaningful emotional attachment. Theoretically, this study expands the understanding of human-agent interaction by identifying these differential pathways. Practically, it suggests that developers must prioritize high visual attractiveness to make VHI interactions acceptable, while simultaneously imbuing agents with human-like qualities to foster lasting emotional loyalty.
Authors
- Mikyoung Kim (ORCID: https://orcid.org/0000-0003-4420-9642)
- Hyun Jung Oh (ORCID: https://orcid.org/0000-0002-1838-7704)
Institutions
- Hanyang University (KR)
- Hongik University (KR)
Publication Details
- Journal
- Cyberpsychology Behavior and Social Networking
- Published
- 2026-09-21
- DOI
- https://doi.org/10.1177/21522715261488664
- Primary Topic
- Evolutionary Psychology and Human Behavior
- Type
- article
- Field-Weighted Citation Impact
- 0.00