Artificial Intelligence for Automated and Semi-Automated Orthodontic Appliance Design in 3D Printing Workflows: A Narrative Review

Abstract Artificial intelligence (AI) and additive manufacturing are increasingly integrated into digital orthodontic workflows, but orthodontic appliance design remains less automated than manufacturing. This narrative review synthesizes the current role of AI-related methods in automated and semi-automated orthodontic appliance geometry design within three-dimensional printing workflows. PubMed, Scopus, and Web of Science were searched for publications from 2015 to 2026, with selected foundational sources included. Current AI-related methods support segmentation, anatomical recognition, surface alignment, treatment-planning support, monitoring, and computer-aided design refinement rather than autonomous geometry generation. Higher automation is most evident in standardized appliances, particularly occlusal splints, clear aligners, and selected indirect bonding tray workflows, whereas fixed functional appliances, anchorage devices, TAD guides, and individualized wire-based constructions remain largely clinician-driven. Key barriers include anatomical variability, limited biomechanical-response prediction, limited explainability, regulatory uncertainty, and lack of appliance-specific datasets. Future progress will likely depend on clinician-supervised hybrid workflows integrating computational design support with biomechanical validation.