Chronic diseases such as diabetes pose significant challenges to healthcare systems due to their long-term impact and potential complications. This study employs machine learning techniques, specifically Gradient Boosting methods, to develop a classifier for predicting the onset of complications in diabetic patients using administrative healthcare data. In addition to a global analysis, we investigate, through Explainable AI, how sex–gender differences influence both the development of complications and the healthcare management of patients. This study highlights the importance of personalized predictive modeling and the necessity of integrating a sex–gender perspective in chronic disease management, emphasizing the role of population-based analysis in uncovering disparities in treatment and healthcare outcomes.