Artificial intelligence (AI) has seen rapid advancements, establishing itself as a robust tool for fostering accessible teaching and learning environments. Nevertheless, the efficacy of these intelligent learning settings in bolstering the engagement and learning outcomes for students with hearing impairments warrants further investigation. This research leveraged Design-Based Research (DBR) to iteratively refine intelligent learning environments, while also adopting the Community of Inquiry (COI) and Technology Acceptance Model (TAM) to assess classroom presence and perception among students with hearing impairments. The results revealed notable enhancements in teaching presence, cognitive presence, social presence, and overall perception within intelligent learning contexts for hearing-impaired students. Additionally, their understanding of abstract and ill structured knowledge significantly improved. Such findings offer valuable insights for the design of inclusive teaching and learning environments in special education.