This article deeply explores the ethical risks caused by the widespread application of AIGC technology in the process of intelligent development of higher education, and is committed to strengthening ethical leadership to achieve effective control of these ethical risks. The first step in building an ethical risk control system is to establish evaluation indicators and a complete system for the ethical risk control mechanism of AIGC technology integrated into university education scenarios. This system is based on ethical principles as its core architecture, with evaluation indicators and specific methods as the supporting foundation. The main content is constructed around the three key steps of risk identification, risk analysis, and risk evaluation. At the same time, attention should be paid to the close connection between the subject and object, and effective linkage should be achieved through prevention and control governance mechanisms to ensure the orderly operation of the entire system. At the level of technical optimization, this article innovatively proposes the use of an improved differential evolution algorithm to optimize RBF neural networks. Specifically, in the process of improving the differential evolution algorithm, a dual archive population optimization strategy was introduced. By creating two different files to record the individual information that was eliminated during the algorithm running process, and applying this information reasonably to the next generation of update iterations. The implementation of this strategy can effectively maintain the novelty and diversity of the population, significantly improve the overall performance of the algorithm, and thereby enhance the accuracy of RBF neural network in predicting ethical risks. Through practical case analysis and verification, an ethical risk assessment index system and grading standards were first constructed for the application of AIGC technology in university education scenarios. Subsequently, the optimized model was used for training and testing, and its evaluation results were compared and analyzed with traditional models. The results showed that the optimized model demonstrated higher reliability and accuracy in evaluating the ethical risks of AIGC technology applied in university education. The series of methods and strategies proposed in this article are of great significance for enhancing the ability to control ethical risks in the application of AIGC technology in university education. They can provide strong theoretical support and practical guidance for the effective prevention and control of ethical risks in the process of intelligent transformation of higher education.
