Stigma remains a major barrier to equitable health and well-being, while artificial intelligence (AI) is increasingly recognized as a tool with both potential and risk in addressing this challenge. However, research on AI and stigma is fragmented across disciplines, hindering a unified understanding of their intersection. To consolidate existing evidence, we conducted a scoping review of 11,769 records published between 2016 and 2025 and identified 70 studies examining the relationship between AI and health-related stigma. Four research themes emerged: AI measuring stigma (n = 42, 60%), stigma influencing AI use (n = 15, 21%), AI increasing stigma (n = 9, 13%), and AI reducing stigma (n = 4, 6%). Most studies focused on mental health disorders, revealing an imbalance in attention to other health conditions. Across studies, we observed inconsistent definitions of stigma, limited cross-cultural perspectives, and few evaluations of real-world AI applications. Addressing these gaps will be critical for developing responsible and equitable AI systems that mitigate rather than reinforce health stigma across broader societal and health contexts.
Song et al. (Sat,) studied this question.