Illustrating Safety: The Role of Digital Media and AI in Designing Inclusive Protection Stories for children with Autism Spectrum Disorders

Abstract

This study investigates the development of an inclusive children’s storybook on self-protection through the integration of digital media and artificial intelligence (AI). The research employed a design-based approach combining AI-assisted creative development with small-group classroom testing. In the development phase, large language models and image-generation tools were used to generate narrative ideas, character concepts, and preliminary illustration sketches, which were subsequently refined through human illustration and pedagogical review. The testing phase involved an interactive read-aloud session with ten students with intellectual learning difficulties at a special education school in Indonesia. Data were collected through structured classroom observation focusing on student attention, verbal participation, non-verbal interaction with illustrations, emotional responses, and recognition of repeated narrative elements. The results indicate that the storybook was able to sustain student attention, encourage moderate levels of verbal participation, and generate generally positive emotional responses during storytelling activities. Students also demonstrated recognition of repeated narrative patterns, suggesting that simplified narrative structures and visual clarity may support engagement and comprehension among learners with diverse cognitive abilities. The findings highlight the potential of AI-assisted design workflows to accelerate the creation of inclusive educational materials while emphasising the continued importance of human oversight to ensure cultural sensitivity, accessibility, and pedagogical relevance..