Abstract: Generative artificial intelligence deeply interacts with the network through technologies such as dynamic paths of multi-modal preset architectures， multi-dimensional group relative strategy optimization reasoning mechanisms， and linkable multi-layer data verification frameworks， thus generating human-like language mechanisms and thinking logic. This provides new opportunities for the digital and intelligent development of college students’ ideological and political education. However， its application faces risks such as the imbalance of ethical relationships under the disembodiment of education， the weakening of ethical governance under the effectiveness of education， and the inappropriate ethical behaviors under the digital and intelligent education， mainly for the following factors： the technological dependence leads to the “de-subjectification” of educational practice activities， the cooperation between the capital logic and the technological logic obscures the essence of educational practice activities， and the weak ethical literacy of artificial intelligence leads to the aggravation of algorithmic risks. Therefore， it is necessary to strengthen the educational subjectivity of technological applications， build a coordinated ethical relationship， establish a correct technological axiology， formulate an intelligent ethical norm system， improve the ethical literacy of the subject in artificial intelligence， and implement appropriate ethical behaviors， so as to help generative artificial intelligence provide strong impetus for the digital and intelligent development of college students’ ideological and political education.