This study aims to explore the application of educational technology in creating picture books for special education and analyze its impact on undergraduate students’ understanding of special education theories. Fifty post-baccalaureate students participated, using ChatGPT to generate picture book texts and tools like Canva and Bing Image Creator to design illustrations. The latest version of the picture books was used for shared reading activities in preschool, and feedback from children was incorporated to revise the content. The results show that students created picture books covering six types of special needs, such as hearing impairment, autism, and ADHD. Peer reviews indicated that about one-third to half of the group works met evaluation standards, and students generally agreed that creating picture books helped deepen their understanding of special education theories. While ChatGPT assisted with text generation, multiple revisions were needed to tailor content to children’s needs, highlighting the crucial role of professionals in adapting language and shaping scenarios. Children’s feedback helped students refine their picture books, aligning them better with developmental characteristics and enhancing educational effectiveness. Through this process, students learned about emerging AI technologies and digital tools, boosting their confidence and sense of achievement in innovative teaching and creation.