Integrative Analytical Framework: A Literature Review of the Ethical and Social Challenges of AI in Healthcare and Rehabilitation

This study aims to synthesize and analyze the ethical and social challenges of applying artificial intelligence (AI) and machine learning (ML) in healthcare and rehabilitation, with a focus on identifying gaps in the literature and opportunities to develop new analytical frameworks. The method used is a systematic literature review based on narrative synthesis of 27 international journal articles published between 2020 and 2025. The results reveal four key interrelated challenges: algorithmic bias originating from data and systems, privacy risks that threaten autonomy, Lack of model transparency, and accountability and governance gaps. Furthermore, the study found a significant gap between the many proposed mitigation strategies (technical, normative, and institutional) and the limited number of evaluation reports on their effectiveness in clinical practice. It concludes that the discourse on AI health ethics remains dominated by conceptual and fragmented approaches, necessitating a shift towards empirical-evaluative research and the development of a more integrative, contextually grounded analytical framework to guide responsible implementation