Abstract
BACKGROUND: Although children with high functioning autism have good mathematical logic potential, the limitations of traditional teaching rhythm and feedback can hinder classroom adaptation and emotional regulation. Artificial Intelligence (AI) tutoring systems provide personalized learning through adaptive approaches and real-time feedback, while existing research mainly focuses on academic achievements and has limited attention to emotional impact. Based on this, the study examined the learning effectiveness and emotional outcomes of AI based mathematics tutoring systems in this group. METHODS: The study adopted a quasi experimental pre and post control design, selecting 60 children with high functioning autism diagnosed by professional institutions as the research subjects, and randomly dividing them into an experimental group and a control group of 30 each. The experimental group used an AI math tutoring system on the basis of routine teaching, while the control group only received routine teaching. The baseline characteristics of the two groups were comparable. The intervention period is 10 weeks, and the system should be used no less than 3 times a week. The AI systems adaptively adjust learning content based on learning behavior data and provide positive feedback. The study evaluated math scores, learning anxiety, emotional stability, and learning engagement using standard chemical tests and teacher and parent rating scales before and after intervention. The intervention effect was analyzed using paired sample t-test and inter group comparison, with a significance level set at p<.05. RESULTS: The results showed that after intervention, high functioning children with autism showed significant improvements in learning and emotion related indicators. Mathematics scores significantly improved after intervention, while learning anxiety scores significantly decreased, indicating that AI mathematics tutoring systems help alleviate negative emotions related to learning while promoting academic performance. At the same time, children's emotional stability and learning engagement scores significantly increased, indicating that the intervention not only improved learning efficiency, but also improved emotional regulation and participation in the learning process. The differences in various indicators before and after intervention were highly statistically significant (p<.001), and the overall results support the positive role of AI mathematics tutoring system in learning support and emotional improvement of high functioning children with autism. DISCUSSION: The results indicate that the AI math tutoring system not only effectively improves the math learning performance of high functioning children with autism, but also plays a positive role in alleviating learning anxiety, improving emotional stability, and enhancing learning engagement. In practice, integrating AI adaptive learning systems into special education curricula can help create inclusive, low stress, and personalized learning environments, providing new paths for the long-term learning and psychological development of high functioning children with autism.