BACKGROUND: The integration of artificial intelligence (AI) into clinical dentistry has opened new avenues for diagnostic support and treatment planning. Among AI tools, large language models such as ChatGPT have shown potential in assisting clinical reasoning. Prosthodontics, due to its interdisciplinary complexity and individualized treatment protocols, presents a particularly demanding context for evaluating such technologies. This systematic review critically examines the clinical performance of ChatGPT in prosthodontics, focusing on diagnostic accuracy, treatment planning, and decision-making. METHODS: A systematic search was conducted across PubMed, Scopus, Web of Science, and the Cochrane Library for studies published between January 2022 and August 2025. Inclusion criteria encompassed original research assessing ChatGPT's performance in prosthodontic contexts. Eight studies met the eligibility criteria, including observational, experimental, and case-based designs. Data extraction and synthesis adhered to ENTREQ and PRISMA guidelines to ensure methodological transparency and reproducibility. RESULTS: ChatGPT-4 consistently outperformed earlier versions in terms of coherence, clinical relevance, and reasoning quality. Applications ranged from implant-supported prosthesis planning and occlusal analysis to prosthetic shade selection. Structured prompting techniques, particularly Chain-of-Thought, significantly enhanced response accuracy and consistency. Across the eight included studies, most reported moderate to high response reliability, although variability remained in complex clinical scenarios. Repeatability analyses revealed moderate to high intra-model reliability, although susceptibility to hallucinations and variability in complex scenarios remain limitations. Most studies were simulation-based, with limited validation in real clinical settings. CONCLUSIONS: ChatGPT exhibits promising capabilities as a supplementary tool in prosthodontics, particularly for educational purposes and preliminary clinical reasoning. However, its direct clinical application should be approached with caution and under professional oversight. Future research must prioritize empirical validation, standardized evaluation protocols, and regulatory frameworks to ensure safe and effective integration into prosthodontic practice.
García et al. (Fri,) studied this question.