Adaptive Artificial Intelligence for Students with Specific Learning Disabilities in Reading Science Content

Abstract The growing integration of generative artificial intelligence (AI) technologies, including systems such as ChatGPT, into educational environments in science presents new opportunities to support learning. However, mainstream AI tools often fail to adequately assist students with specific learning disabilities in reading, such as dyslexia. Students with reading disabilities often require specialized instruction tailored to the unique challenges posed by difficulties in reading comprehension, decoding, and retaining multi-step directions often present in complex science texts. While current AI technologies can provide basic explanations, they lack real-time, adaptive guidance with step-by-step feedback personalized to individual science learners. Additionally, predominantly text-based AI often does not suit the needs of students who benefit from interactive, multimodal learning strategies including visual aids. To better serve the needs of neurodiverse learners in science classrooms, AI systems must evolve with a focus on inclusivity. Potential improvements include adaptive learning algorithms based upon the use of neurological data, enhanced formative assessment and feedback techniques, and incorporation of graphics and other multisensory features. With innovative designs that align to principles of universal learning, AI-based educational technologies could support personalized and individualized reading skill development for all students. This will require sustained efforts to develop AI that is responsive to diverse learning needs.