In semi-arid inland basins, groundwater resources are already scarce, and the presence of high fluoride (F-) concentrations in groundwater poses potential health risks to residents, further compromising the safety of drinking water. Seventy groundwater samples were collected from the ChahanNur Basin to investigate the factors controlling high F- concentrations. Hydrogeochemical analysis combined with machine learning techniques was applied to identify the key processes driving fluoride enrichment in the region. The results revealed that high F- groundwater is primarily concentrated in the western-central part of the basin, where the predominant chemical types are Cl-Na and Cl-Ca. Regression analysis of water chemistry indicated that F- enrichment was influenced by hydrogeochemical processes, including the dissolution and precipitation of fluoride-bearing minerals, competitive adsorption, cation exchange, and evaporation. The CatBoost model, utilizing these four key factors as predictor variables, was applied to predict groundwater F- concentrations. The model's performance was evaluated by calculating the root mean square error (RMSE) for both the train and test sets, which yielded values of 0.18 and 0.22, respectively. The relative contributions of mineral dissolution and precipitation, evaporation, competitive adsorption, and cation exchange to groundwater F- enrichment were 66.2%, 18.3%, 11.2%, and 4.3%, respectively. Furthermore, the risk assessment identified regions where different population groups are particularly affected by high F- concentrations in groundwater. These findings provide a scientific basis for the effective management of high F- groundwater in inland basins, ensuring the safety of water supplies for local communities.