This article by Gripp et al examines how artificial intelligence (AI) is used as a supportive tool of facial feature analysis. It is being used to improve the diagnosis of genetic and syndromic conditions within pediatric medicine. The research synthesizes findings across studies using 2D image-based tools such as D-score, DeepGestalt, and GestaltMatcher, leading to a review of how these algorithms perform when integrated with clinical features. The review found that tools like D-score can accurately distinguish between typical and syndromic facial features. DeepGestalt can effectively rank potential syndromes based on facial similarity, and GestaltMatcher enables the identification of ultra-rare disorders by comparing single patient images. They also emphasized the importance of combining facial analysis with human phenotype ontology (HPO) terms and next-generation sequencing to improve variant prioritization and diagnostic accuracy. Databases like the GestaltMatcher Database (GMDB) and generative AI tools such as GestaltGAN can help benchmark machine learning performance and expand research by providing diverse reference images without privacy concerns. The author emphasized that these tools can enhance clinicians’ ability to recognize subtle or rare dysmorphic features and will likely become increasingly integrated into electronic medical records and routine variant analysis pipelines. The article also stressed that training models on ethnically diverse images is essential for accuracy and equitable use.