Complex and dynamic interactions between environmental conditions, pollution layers, and hydrophobic surfaces under high-voltage stress remain poorly understood. Artificial Intelligence and Machine Learning provide the opportunity to overcome these limitations by discovering hidden relations between leakage current activity and meteorological variables.