The concept of digitalization and intelligence is constantly emerging with the rapid development of information technology, and the in-depth integration of digital intelligence empowerment and teacher education development has garnered significant attention. In this paper, we first examine the stage of teacher development under digital empowerment, tracing the path of teacher education development. We then construct the digital and intelligent TPACK model, as well as the model of teacher education development ability. Subsequently, we adopt a BP neural network to evaluate teacher education development and optimize it using a genetic algorithm. After testing the model, we develop a teacher education development evaluation system based on the GA-BP algorithm, which we then apply and analyze. This paper’s optimized model reduced absolute and relative errors by 418.281 and 3.53%, respectively, compared to the traditional model, and improved the overall goodness of fit by 0.0144. Following the implementation of the system in this paper, the teacher assessment analysis revealed an average teaching ability score of 89.75 for teachers majoring in Japanese foreign language, and 90.26 for those majoring in English foreign language. Notably, the average level of English foreign language majors and their various dimension levels were both higher. Secondly, foreign language teachers with two years of training have an average teaching ability score of 91.32, which is 5.07 points higher than those with only one year of training. It proves the effectiveness of the practical application of this system, and this paper also provides suggestions for the development of teacher education by both teachers and schools, which is of practical significance for the optimization and improvement of foreign language teachers’ education in colleges and universities.