The impact of generative artificial intelligence on the social-emotional skills of children with autism spectrum disorder: a meta-analysis of 38 experimental and quasi-experimental studies

The impact of Generative Artificial Intelligence (GAI) on the social-emotional skills of children with Autism Spectrum Disorder (ASD) has yielded inconsistent conclusions in academic research. This study employs a meta-analysis approach to synthesize empirical findings and assess the effects of GAI on the development of social-emotional skills in children with ASD. A comprehensive literature search was conducted using multiple academic databases, covering studies published between 2014 and 2024. The search process involved keywords related to GAI, ASD, autis*children, and social-emotional et al. with a focus on empirical studies that utilized controlled experimental designs. A total of 38 studies were included in the meta-analysis, which was performed using the CMA 3.0 system. The results indicate that GAI has a significant overall positive effect on the development of social-emotional skills in children with ASD. Additionally, GAI demonstrates a significant positive impact on all five dimensions of social-emotional skills, with no significant difference between groups. Three moderating variables—experimental duration, sample size, and type of GAI—were identified as key factors enhancing the social-emotional skills of children with ASD, with significant differences in inter-group effects. Based on these findings, it is recommended that future research optimize contextual support, reconsider the role of teachers, focus on the type of technology, and tailor interventions to the scale of special needs to maximize the effectiveness of GAI in promoting social-emotional development in children with ASD.