 This study proposes a novel methodology for detecting leaks in water carrying pipelines. It uses a graph convolutional neural network, GCN, to analyze the network structure of pipelines and then applies a machine learning algorithm to identify anomalous behavior. The GCN model outperformed a support vector machine SVM model in terms of accuracy, demonstrating its effectiveness in detecting leaks in pipeline systems. The findings have important implications for water resource management and environmental protection. This article was authored by Erson Sahin and Hussein Yousse.