It has been studied to analyze the operations in nuclear power plants (NPPs) where the human error could make a trigger to the nuclear disaster following the earthquake in South Korea. The earthquake warning system and its related control algorithm are modeled by the machine learning (ML) of artificial intelligence (AI) where the neural networking is done as a major role. The AI-based control system could give the very good performance in the earthquake incident which is shown by the system dynamics (SD) based simulations. Hence, the nuclear safety system needs the AI based alarming technology which could be a much more advanced control system incorporated with the conventional multiple-barrier concept in NPPs.