Artificial intelligence (AI) and machine learning (ML) have emerged as important technologies in modern environmental monitoring as they enable researchers and institutions to analyze huge, complex and continuous environmental datasets. This paper reviews the development of AI-based environmental monitoring up to December 2025, based on recent bibliometric evidence, with a focus on air quality, water quality, climate modelling, biodiversity monitoring, disaster prediction, remote sensing, Internet of Things (IoT) systems and explainable AI. The assessment notes that AI has advanced from experimental modelling to operational decision-support systems, integrating satellite data, ground sensors, meteorological records, camera traps, drones, and real-time IoT networks. However, key remaining issues include uneven data availability, black-box model behavior, insufficient validation in low-resource locations, cybersecurity threats in IoT systems and limited policy integration. The study makes the case that the next phase of AI-based environmental monitoring should emphasize explainability, data fairness, interdisciplinary collaboration and human-centered environmental governance.