<p>Bioactive glasses bond with biological tissues, promoting tissue regeneration. Their chemical constitution directly influences bioactivity, mechanical strength, biocompatibility, and degradation kinetics. However, because composition–property relationships in multicomponent bioactive glasses are highly non-linear and often involve complex interactions among constituents, accurate prediction remains a challenging task. Traditional approaches rely on expensive, iterative empirical research, time-consuming interpretation of results, as well as computer simulations. To address these challenges, AI and ML offer transformative predictive analysis that can accelerate materials discovery. This review highlights the successful application of AI-based models, ranging from Gaussian process regression and random forests to deep neural networks, in reliably predicting the physical, mechanical, chemical, structural, and biological properties of different bioactive glass systems based on their composition and enhancing glass formulations for various applications. To provide a comprehensive assessment of the field, the underlying datasets, feature selection techniques, model performance measurements, uncertainty quantification strategies, and explainable AI approaches are examined. Despite the growing use of AI/ML techniques, standard datasets and model interpretability remain lacking. Addressing these problems will enable AI to be fully employed in bioactive glass research, thereby expediting the creation of new compositions and enhancing biomaterial quality, effectiveness, and modification for regenerative treatments, ultimately improving good health and well-being as well as patient outcomes.</p> Graphical abstract <p></p>