Abstract This article explores the ethical and cultural dimensions of virtual influencers (VIs), computer-generated characters used in emotionally persuasive digital environments, through a cross-cultural qualitative study conducted in Myanmar, South Korea, Singapore, and the United States. While VIs are rapidly gaining traction in advertising and social media, they raise underexamined questions about emotional authenticity, trust, and the ethical boundaries of simulated social interaction. Drawing on focus group discussions with 19 participants, including nationals and residents with varied degrees of familiarity with VIs, the study applies three theoretical frameworks: the CASA paradigm (Computers Are Social Actors), the Uncanny Valley hypothesis, and Diffusion of Innovation theory. Participants were shown curated examples of real-world virtual influencers and invited to reflect on issues of realism, machine identity, emotional engagement, and ethical discomfort. Findings reveal that responses to VIs are not universal but shaped by cultural norms, levels of digital literacy, and prevailing media discourses. Participants across contexts expressed distinct thresholds for emotional resonance and divergent ethical concerns, particularly regarding disclosure, manipulation, and perceived authenticity. The article contributes to debates in AI ethics by introducing three conceptual lenses, synthetic misrecognition, emotional transparency, and meta-authenticity, and argues for culturally responsive frameworks in the governance, labelling, and design of emotionally expressive synthetic personas. It concludes by proposing context-sensitive strategies for platform design, media literacy, and public policy, and calls for a broader societal dialogue around emotional AI and the future of human–machine intimacy.
Khin et al. (Sat,) studied this question.