CULTURAL ADAPTATION IN AVATARS AND VIRTUAL AGENTS: A SYSTEMATIC LITERATURE REVIEW

Abstract

Culturally adapted virtual avatars are presented as a way to increase user identification with the tool, though little is known about the actual reach of that effect. Reported according to PRISMA 2020 guidance, this systematic review analyses 29 studies published between 2020 and 2026 and retrieved from Scopus, IEEE Xplore, and the ACM Digital Library on culturally adapted avatars and virtual agents. Findings were organised by uses, adaptation components, benefits, challenges, and evidence of impact. The findings suggest that effectiveness is associated with coherence among language, appearance, behaviour, task, and context, and that isolated cultural cues produce different effects across audiences. Some studies associate adaptation with greater trust, engagement, satisfaction, or identification, but the evidence remains heterogeneous and often relies on small samples, brief exposure, and acceptance measures. The field still lacks a unified conceptual model and standardised cross-cultural metrics, and author affiliations are concentrated in North America, Europe, and parts of Asia. The synthesis indicates that community participation, validation with target users, and explicit reporting of limitations emerge as relevant practices for avoiding stereotypes, improving methodological transparency, and guiding the design of more inclusive agents across culturally diverse contexts.