A model training method is disclosed. The model training method includes separately inputting extracted multimedia data into a teacher network and a student network. The student network is trained based on a difference between a feature output by the teacher network and a feature output by the student network. The student network can output a feature that is similar to an output feature of the teacher network when processing non-AI-generated multimedia data. The student network can output a feature that is greatly different from an output feature of the teacher network when processing AI-generated multimedia data.