The Inclusive Algorithm: A Systematic Review of AI and Machine Learning in Supporting Learners with Disabilities

Inclusive education has emerged as a global priority, emphasizing equitable access, participation, and learning outcomes for learners with disabilities. Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) have introduced new opportunities to address diverse learner needs through adaptive, personalized, and accessible educational technologies. This systematic review synthesizes empirical evidence on the effectiveness of AI- and ML-based interventions for learners with disabilities across educational contexts. Guided by PRISMA 2020 standards, a comprehensive literature search was conducted across Scopus, Web of Science, ERIC, IEEE Xplore, and Google Scholar, identifying 245 peer-reviewed studies published between 2015 and December 2025. Following duplicate removal, screening, eligibility assessment, and quality appraisal, 19 studies met all methodological and thematic inclusion criteria and were included in the final thematic narrative synthesis. The review examined types of AI/ML technologies, disability categories (learning, sensory, physical, and psychosocial), educational and inclusion-related outcomes, and ethical and accessibility considerations. The included studies employed quantitative (47.4%), qualitative (31.6%), and mixed-methods (21.0%) designs. AI-driven interventions, such as intelligent tutoring systems, natural language processing applications, assistive technologies, and learning analytics, demonstrated positive effects on academic achievement, accessibility, learner autonomy, engagement, psychosocial outcomes, and social inclusion, with particularly strong evidence for learners with learning and sensory disabilities. However, evidence for institutional-level impact and long-term outcomes remains limited. Key challenges identified include algorithmic bias, data privacy risks, uneven accessibility compliance, and persistent inequities between high-income and low-resource contexts. Overall, the findings indicate that AI and ML can meaningfully support inclusive education when grounded in Universal Design for Learning (UDL) principles and rights-based frameworks. The review underscores the need for more methodologically rigorous, geographically diverse, and longitudinal research to determine which technologies are most effective for specific disability groups and to ensure that AI-enabled education advances inclusion rather than reinforcing existing inequities.