Osteoporosis is a disease that causes bones to become weak and brittle, increasing the risk of fractures. Research using machine learning (ML) techniques has been able to make progress in understanding and managing osteoporosis, but there are technical and clinical concerns about the use of ML in this field. A review of 89 studies on the use of ML for various aspects of osteoporosis management found that the quality of the studies was generally moderate, with common limitations including incomplete reporting, inadequate data splitting, and a lack of external validation. However, the use of images for opportunistic osteoporosis diagnosis or fracture detection appeared to be a promising approach, and research on using ML to identify new fracture risk factors and improve fracture prediction was also promising. It is suggested that the use of standardized checklists for developing and sharing ML models could help to address some of the limitations identified in the reviewed studies.