Artificial Intelligence versus human evaluation of pediatric smile aesthetics: a comparative study across multiple observer groups

Aims: The aim of this study was to comparatively evaluate how pediatric smile aesthetics are perceived by children, parents, dentists, and Artificial Intelligence (AI) systems, with a particular focus on the potential role of AI in clinical decision support. Methods: The study utilized nine standardized frontal smile photographs featuring different aesthetic modifications of the maxillary central incisors. A total of 354 human participants (87 children, 97 parents, 80 pediatric dentists, and 90 general dentists) evaluated the images using binary aesthetic preference questions and a 0-10 aesthetic rating scale. Additionally, four AI systems (ChatGPT-5.2, Gemini 3, Grok 4.1, and Microsoft Copilot) independently evaluated the same images. Categorical data were analyzed using Chi-square-based tests, while aesthetic scores were analyzed using a generalized linear model (gamma distribution, log link function). Statistically significant differences were found between evaluator type and aesthetic preferences for most questions (p<0.05). Results: When all groups were evaluated together, the highest aesthetic scores were assigned to photo I, representing preserved anterior integrity, whereas the lowest scores were assigned to photo G, reflecting caries involvement. In the generalized linear model analysis, evaluator type, photograph, and evaluator×photograph interaction were found to be significant (p<0.001). AI systems tended to assign higher and more homogeneous aesthetic scores than human evaluators, suggesting a more standardized but potentially less context-sensitive assessment pattern. Within the limitations of this cross-sectional study, the perception of pediatric smile aesthetics varies depending on evaluator type and visual characteristics. Conclusion: The observed differences between AI and human evaluators suggest that AI systems may influence aesthetic assessment patterns, with potential implications for AI-assisted clinical decision-making and patient communication. However, AI systems should be considered as supportive tools and cannot replace clinical judgment, particularly in the context of pediatric dentistry where developmental and clinical factors must be carefully considered.