AI FOR DISABILITY SUPPORT: A SECURE FRAMEWORK USING GENERATIVE MODELS, RL, AND FL

Artificial Intelligence (AI) is revolutionizing personalized healthcare by offering promising solutions for individuals with disabilities. However, persistent challenges remain particularly in ensuring data privacy, real time adaptability,and inclusivity.This review explores how combining three AI paradigms Generative AI, Reinforcement Learning (RL), and Federated Learning (FL) can address these limitations. Through thematic analysis of over 50 peer-reviewed studies published between 2018 and 2024, we identify the unique and synergistic contributions of these technologies in enhancing healthcare delivery for disabled populations. We propose a novel, secure, and adaptive framework that integrates:Generative AI for inclusive multimodal interfaces and synthetic health data generation Reinforcement Learning to enable real-time system adaptation based on user interaction Federated Learning to ensure privacy-preserving, decentralized data processing The framework is illustrated with practical applications in mobility, sensory,and cognitive support.This review aims to guide future research toward building AI driven healthcare systems that are secure, inclusive, and responsive to the diverse needs of the disabled community.