This research looked at how using AI in education affects long-lasting learning results and maintaining English language abilities and vocabulary for Iranian intermediate EFL learners. The study used a quasi-experimental design. 60 students from a school in Tehran were split into two groups. The experimental group used AI tools like adaptive platforms, personalized feedback, and simulations. The control group learned with traditional teaching methods. Both groups took pre-tests, went through 12 weekly learning modules over four months, completed post-tests and then had a retention test six months later to see how well they understood English grammar, picked up new vocabulary, and developed language skills. The results showed that the AI group had a much higher average score and better consistency in keeping the language learned. The regression analysis also showed that using AI was a strong reason for improvement in performance. Feedback from focus groups revealed that students were more involved because of the personalized learning and instant feedback. About 78% of the participants in the AI group said they felt more motivated. These findings suggest that AI-assisted education is effective for achieving long-term academic success, especially through tailored content and interactive problem-solving tasks that help retain English skills over time. The study recommends that schools embrace AI in their curricula to increase the long-lasting retention of language skills and critical thinking in intermediate pupils.
