Based on the development and application of intelligent technology, educational institutions are facing multiple risks and challenges while ushering in significant development opportunities. In order to cope with the increasingly complex and diversified risk problems and realize the high-quality development of educational institutions, this paper, based on the evaluation principle of objectivity and risk avoidance, identifies the risks of the educational application of artificial intelligence technology by applying the Delphi method. Using the hierarchical analysis method to screen the index system, we established a risk assessment index system for the educational application of artificial intelligence in educational institutions. Analyze the degree of risk impact and combine the risk level with Borda value to obtain the educational risk assessment results. According to the risk assessment results of educational institutions, risk avoidance strategies in theory, system, and construction are proposed. Overall, the factors with higher or lower educational risk severity rankings of AI technology in educational institutions account for a relatively small percentage, and 70% of the risk factor items have a risk severity level between I and III.