LIMITAÇÕES DA INTELIGÊNCIA ARTIFICIAL NO PLANEJAMENTO ESTÉTICO ODONTOLÓGICO: UMA REVISÃO DE ESCOPO E PROPOSTA CRANIOMÉTRICA INDIVIDUALIZADA PARA DEFINIÇÃO DA LARGURA DO INCISIVO CENTRAL SUPERIOR

Objective: To map and analyze the available scientific evidence on the limitations of Artificial Intelligence in dental aesthetic planning, with emphasis on the definition of the width of the maxillary central incisor. Materials and methods: This is a scoping review conducted in accordance with the guidelines of the Joanna Briggs Institute. Studies published between 2022 and 2025 that addressed the use of digital technologies and Artificial Intelligence applied to dental aesthetic planning were included. The selection included methodological, observational, clinical and diagnostic studies, with no language restrictions. The data were extracted, organized and synthesized in a descriptive way, considering methodological, technological, aesthetic characteristics and the methodological quality of the studies. Results: Ten studies were included, predominantly developed in China, with samples ranging from large databases to clinical studies with a small number of cases. Wide adoption of digital technologies was observed, with recurrent use of Artificial Intelligence associated with CAD-CAM flows, Digital Smile Design, dental segmentation, radiographic diagnosis and automated design of restorations. Despite the satisfactory technical performance and the low risk of methodological bias, the studies showed limitations related to aesthetic individualization, with a predominance of the use of standardized models and population means, especially in the definition of the width of the maxillary central incisor. Final considerations: Artificial Intelligence has high potential to optimize dental aesthetic planning; however, its limitations regarding personalization reinforce the need to incorporate individualized craniometric parameters and clinical judgment into algorithms, aiming at a more accurate, humanized and evidence-based planning.