The electric arc furnace accounts for approximately one-third of global steel production. Even in scrap-based electric arc furnace steelmaking, achieving the target phosphorous content in the final steel product poses a challenge. Enhancing dephosphorization efficiency requires consideration of various initial conditions and operational parameters, rendering process improvement both scientifically and technically demanding. In this study, an artificial neural network model was employed to predict the end-point phosphorus content of steel. A dataset comprising over 1760 entries with 12 input parameters was collected from the plant. Rigorous data processing techniques were applied to enhance both dataset quality and model performance. The findings demonstrate that the model can accurately predict the final phosphorus content of the steel, at least within the tested operational conditions.