Basketball tactics teaching occupies a major position in both professional sports teams and schools, even in social basketball training institutions. The paper suggests an improvement plan based on the shortcomings of the generative AI grand model, and integrates the enhanced model into the basketball tactics teaching process to create a visualization method for basketball tactics evolution. The proposed teaching model was applied to students’ basketball tactics training courses, and the use of the teaching experiment was investigated by using comprehensive research methods such as questionnaire survey method, paired-sample t-test, and one-way ANOVA, with students majoring in physical education in GT College as the experimental subjects. The results showed that the individual offensive and defensive tactics, local offensive and defensive tactics, and team-wide offensive and defensive tactics of the students in the experimental group and the control group were significantly improved after the experiment relative to the pre-experimental period and that the improvement of the experimental group (9.655-13.989) was more significant than that of the control group (4.844-6.515), and that the experimental group assisted with the basketball tactics teaching model based on the generative AI was more effective than the control group assisted with traditional tactics training method. Control the group using the conventional tactical training method. The study provides a reference for the implementation of teaching reform and improvement of teaching quality in the process of teaching basketball tactics in the future.