Heavy metal-contaminated soils are frequently encountered, resulting from improper waste disposal practices and accidental spills, posing a significant threat to public health and the environment. Electrokinetic remediation (EKR) is proven to be effective in remediating heavy metal-contaminated soils, especially in low-permeable clays and heterogeneous soils. EKR involves applying a low electric potential gradient to facilitate contaminant transport—through various mechanisms including electrophoresis, electroosmosis, and electromigration—toward the electrodes for subsequent removal. The overall success of EKR depends on multiple factors, which includes soil type, contaminant nature and concentration, and electric potential, among others. The objective of the current study is to preliminarily assess the effectiveness of machine learning (ML) in predicting the EKR-induced migration and removal of heavy metals in contaminated soils using a comprehensive database derived from past laboratory studies conducted at the University of Illinois Chicago. This database encompassed various variables related to EKR, soil type, and contaminant properties alongside normalized distance from the cathode and the corresponding pH variations and metal concentrations. Four different regression machine learning (ML) models—random forest (RF), gradient boosting (GB), categorical boosting (CatBoost), and artificial neural network (ANN)—were trained and tested using the compiled database to predict pH distribution and metal migration post EKR. Notably, both the RF and GB models effectively predicted pH distribution post EKR. All the models except RF effectively predicted selected metal migration with similar migration patterns. However, none of the models effectively predicted EKR-induced migration for metals that exist in different oxidation states (e.g., chromium) at different pH conditions, or when trained on data containing diverse contaminants with wide-ranging properties. To address this, process-informed AI should be explored in future studies to accurately capture various complex underlying variables and processes to accurately predict the efficiency of EKR.