Artificial Intelligence in Plastic Surgery: Current Status, Limitations, and Future Directions

Abstract Artificial intelligence (AI) has become pervasive in and beyond plastic surgery. Myriad applications exist, and patients and plastic surgeons are increasingly turning to AI for information. This narrative review examines the current scope of AI applications in plastic surgery and highlights challenges and limitations based on current literature. A PubMed search for articles about or using AI in plastic surgery published before September 22, 2025 identified 1866 articles. Letters, commentaries, review articles, surveys, and articles not in the English language were excluded. Titles and abstracts were reviewed and studies classified according to AI modality, plastic surgery application, and subspecialty. Studies were classified under multiple categories, if applicable. This narrowed the results search to 460 qualifying articles, of which 54 involved patient education, 35 plastic surgeon education, 79 clinical decision-making, 62 outcome prediction or risk assessment, 46 clinical outcome assessment, 133 diagnosis, 46 practice management, and 17 research. Study methodologies and AI models varied widely. In terms of the types of AI used, 155 articles utilized large language models, 6 natural language processing, 9 text-to-imaging models, and 299 other machine-learning or deep-learning systems. Large language models were most often used in patient education studies, while machine learning predominated in diagnostic studies. AI spans the breadth of plastic surgery, although the literature is limited by heterogeneity. Plastic surgeons must know the advantages and opportunities provided by AI, while recognizing its limitations, pitfalls, and areas needing improvement. Ethical, safe, and forward-thinking AI in plastic surgery requires a multidisciplinary approach involving plastic surgeons, data scientists, ethicists, legal experts, and policymakers.