Artificial intelligence in osteoarthritis and osteoporosis: implications for management and research

Artificial intelligence (AI) is increasingly influencing clinical care and research in osteoarthritis and osteoporosis. For patients, AI can broaden access to health information, including understanding of disease and treatment options. For clinicians, AI-based systems may assist risk stratification and decision-making, including identifying individuals most likely to benefit from specific therapies. In medical imaging, AI approaches can grade osteoarthritis severity, identify occult pathology (including atypical femoral fractures), and estimate low bone mineral density from non-dedicated examinations. In research, AI can facilitate analysis of complex, high-dimensional datasets (e.g., multi-omics and cell-cell communication) supporting the discovery of biomarkers, candidate therapeutic targets and disease endotypes. Emerging directions include agentic AI systems for end-to-end workflows, adaptation of foundation models pre-trained on large, diverse datasets, and carefully evaluated synthetic data for augmentation and validation/quality-control pipelines. This review examines current and emerging AI applications across osteoarthritis and osteoporosis, and pathways for safe, equitable and effective deployment.