Machine elements frequently operate under variable conditions, resulting in significant variations in interfacial friction across different lubrication regimes. The Stribeck curve is a well-established tool for visualizing frictional behavior under boundary, mixed, and full-film lubrication conditions. While numerical models such as Thermo Plasto-Elastohydrodynamic Lubrication (TPEHL) provide accurate friction predictions, they are computationally demanding. This study investigates the application of Artificial Intelligence (AI) to predict the coefficient of friction in Stribeck curves, utilizing a comprehensive experimental dataset based on polyalphaolefin (PAO) oil RENOLIN UNISYN XT ISO VG 68. Three AI models - Neural Networks, Random Forest, and Support Vector Machine - were evaluated using cross-validation. Statistical analysis via Tukey’s Honestly Significant Difference (HSD) test demonstrated that the Random Forest model achieved superior predictive accuracy compared to the Neural Networks and Support Vector Machine models. Subsequently, the Random Forest model was applied to predict Stribeck curves for PAO RENOLIN UNISYN XT ISO VG 150, a lubricant of similar composition but higher viscosity, confirming its robustness and generalization capability across different lubricants.