As software testing complexity continues to increase, traditional teaching models face significant challenges in cultivating automated testing talents. This study proposes an innovative teaching model that integrates generative AI with automated testing, deeply incorporating AI technology into test scenario design, code generation, and result evaluation processes. Practice shows that this model not only effectively reduces students' learning barriers but also significantly improves test case coverage and code quality. This research provides new insights for software testing curriculum reform and offers valuable reference for cultivating testing professionals adapted to the demands of intelligent development.