Generative pre-training models have brought disruptive changes to the education system by recon-figuring the education space and teaching methods with their highly personalised learning experience and dy-namic learning based on data and feedback.However,with the generalisation of generative pre-training mod-els,especially the rise of ChatGPT,a number of new ethical risks have come to the forefront,including tech-nological risks,content risks,data risks and algorithmic risks.These risks range from the violation of hu-man beings by biased or misleading statements unconsciously produced by the generative pre-training models themselves during the training process,to human beings themselves such as:developers'algorithmic hege-mony,individuals'lack of technological awareness,and teachers'lack of data literacy.To address these ethical risks,ethical risk management can be carried out by strengthening theoretical research and knowl-edge popularisation,developing generative pre-training models in line with China's education model,and promoting structural changes in education patterns to ensure that the application of generative pre-training model technology in education is in line with the ethical norms of education.
维普
万方数据