With global climate change and the increasing of water demand, the global water crisis is becoming more and more serious, so the efficient utilization of water resources has become a hot spot for research. Meanwhile, urban water systems are important sources of carbon emissions, and therefore the study of water-carbon nexus in water resource utilization is particularly necessary in the field of carbon neutral policies. In this study, different artificial intelligence (AI) methods were used to simulate and predict the quantitative variations of water and carbon under long time series, including artificial neural network (ANN) and support vector machine (SVM). Additionally, the optimal emission reduction solutions for efficient water resource utilization were proposed based on different scenarios, which might be encountered in the future. Through simulation analysis, ANN could achieve better performance than SVM in both the training and testing phases. Quantitative changes in water demand and carbon emissions from water systems showed opposite trends over the next decade. Among the various scenario simulation in terms of carbon emissions, the application of the smart water resources management platform and the popularization of water-saving appliances could lead to the greatest increase in the efficiency of water utilization (98.1%). This study uses AI methods to explain the water-carbon nexus in the water-utilizing process, so as to propose targeted solutions to substantially improve the efficiency of water utilization and better realize emission reduction.