Microbially induced calcite precipitation (MICP) is an environmentally beneficial and potentially sustainable technique that can strengthen weak soils. This review provides an overview of the prediction of unconfined compressive strength (UCS) of bio-cemented sand by machine learning. MICP has gained significant attention for its potential in stabilizing sandy soils. However, the process involves complex interactions between biological, chemical, and environmental factors, making it challenging to optimize and scale efficiently. Recently, the application of machine learning (ML) algorithms has emerged as a powerful tool to address these challenges by predicting critical outcomes such as soil strength, porosity, and cementation quality based on input parameters. This review paper explores the state-of-the-art in integrating ML with bio-cemented sand research, summarizing key advances in model development, performance, and data management strategies. It discusses various ML techniques applied in this field, including supervised learning models, and highlights the benefits of reducing experimental time and cost. Additionally, the review identifies current limitations, such as data scarcity and model interpretability, while suggesting future research directions for enhancing model accuracy and generalization. It also explores the application of machine learning algorithms to predict high-performance outcomes. The interdisciplinary collaboration between geotechnical engineering, microbiology, and data science is emphasized as essential for driving innovation in sustainable soil stabilization. This review paper not only enhances understanding of MICP-treated sand but also highlights the potential of machine learning in advancing soil stabilization techniques.