Artificial Intelligence and Machine Learning-Based Groundwater Management: A Systematic Review

Rapid urbanization, climate change, and unsustainable groundwater abstraction pose a growing threat to groundwater resources. This creates an urgent need for effective and reliable management techniques. Artificial intelligence (AI) and machine learning (ML) have become powerful tools for groundwater assessment, prediction, and decision support. This systematic review provides a thorough overview of developments in AI- and ML-based groundwater management using the PRISMA 2020 reporting guidelines. A total of 1,948 records were retrieved from Scopus for the period January 2021 to June 2026. Bibliometric and network analyses were conducted using VOSviewer. After the PRISMA 2020 screening procedure, 60 studies that met the inclusion criteria were included for thematic and qualitative analysis. Bibliometric analysis revealed a marked increase in publications, reaching a peak in 2025. China, the United States, and Iran were the leading contributors to the published literature. The largest number of studies appeared in the Journal of Hydrology, Water, Science of the Total Environment, Groundwater for Sustainable Development, and Water Resources Management. Groundwater recharge estimation, groundwater level forecasting, metaheuristic optimization, groundwater potential mapping using Geographic Information Systems (GIS) and remote sensing, and Bayesian uncertainty quantification were the six primary research areas identified by the topic analysis. When compared to traditional hydrological and statistical models, deep learning techniques, particularly Long Short-Term Memory (LSTM), Graph Neural Networks (GNNs), Transformers, and hybrid machine learning models, demonstrated notable prediction performance. This review also highlights emerging research directions, such as Explainable AI, AutoML, cloud computing, physics-informed machine learning, Internet of Things (IoT)-enabled monitoring, and digital twins, which are anticipated to enhance the operational implementations, reliability, and transparency of AI-driven groundwater management. These findings provided a comprehensive evidence base to guide future research, policy development, and implementation of AI decision-support systems for sustainable groundwater management.