School violence remains a global challenge that affects the safety, emotional well-being, and development of students in educational settings. In this context, Artificial Intelligence (AI) has emerged as a promising tool to detect and prevent violent incidents within schools. This study conducts a detailed analysis of how AI technologies are being applied to identify patterns of school violence, examining the most widely used methodologies and current trends in the field. Through a systematic review of academic research, approaches were identified that employ advanced machine learning models, neural networks, and smart sensors to recognize violent behaviors. The results show that techniques such as Decision Trees (DT), Support Vector Machines (SVM), and Convolutional Neural Networks (CNN) have achieved accuracies greater than 90%, demonstrating their potential for early detection of risk situations. However, significant challenges remain, such as the ability of these models to adapt to different educational contexts and the need to ensure that their implementation is ethical and respectful of students’ privacy. Moreover, the study highlights the importance of a multidisciplinary approach that combines technology, education, and psychology to develop more effective strategies for preventing school violence. With responsible application of AI, it is possible to improve school security and create more inclusive and protective learning environments for all students.