Background: Artificial intelligence (AI) is increasingly investigated for radiographic assessment of peri-implant bone conditions, although radiographic marginal bone loss does not alone establish a clinical diagnosis of peri-implantitis. This narrative review aimed to critically synthesize current evidence on AI-assisted radiographic assessment of peri-implant marginal bone levels and bone loss. Methods: A structured literature search was conducted in PubMed/MEDLINE and Scopus. Eligible original studies applying AI or machine-learning methods to human dental radiographs for peri-implant bone assessment were included in a qualitative synthesis. Results: Seventeen studies were included. Periapical radiographs were the predominant imaging modality, while panoramic radiographs and CBCT were less frequently investigated. AI applications included implant and landmark localization, marginal bone-level measurement and quantification, bone-loss severity and defect-morphology classification, and prediction using imaging-derived features. Reported performance was generally promising, although substantial heterogeneity was observed in datasets, target tasks, reference standards, validation strategies, and performance metrics. Conclusions: Current radiograph-based AI systems show potential for image analysis, quantitative assessment, and clinical decision support. However, radiographic AI performance should not be interpreted as autonomous diagnostic accuracy for peri-implantitis. External validation, longitudinal assessment, and integration with relevant clinical parameters remain priorities for future research.