Many regions in arid and semi-arid lands suffer from water resource scarcity due to climatic constraints combined with a rapid increase in domestic water consumption due to population and industry growth. The increased demand for high-quality and sustainable water assets creates an urgent need to investigate groundwater resources as an alternative solution. In this study, weighted overlay analysis based on an analytical hierarchy process and supported by a random forest machine learning technique was utilized to explore groundwater potential zones in the northern United Arab Emirates (UAE). Toward that end, eight thematic layers were prepared and processed in a geographic information science environment to produce the groundwater potential map (GPM): elevation, slope, precipitation, geology, geomorphology, drainage stream density, major fracture, and lineament density. The resultant GPM was classified in five potential groundwater zones ranging from very high to very low. The GPM was validated based on existing well distribution and selected wellfield yielding production. It was found that 81.7% of existing wells are located in the moderate-high zone and higher. Single parameter sensitivity analysis technique revealed that the geology, precipitation, and geomorphology were the most significant factors, with a mean effective weight of 42.7%, 34.22%, and 32.23%, respectively. This research provides substantial insights for local water managers and authorities by providing a GPM as well as geo-hydrological findings in order to achieve sustainable groundwater management in the UAE.