Governing High-Risk AI in Healthcare: Aligning Technical Robustness with Ethical and Legal Accountability

Abstract

Clinical decision-making, diagnostics, and patient management have been transformed by the integration of high-risk artificial intelligence (AI) systems in healthcare, and complex technical, ethical, and legal issues have been generated simultaneously. These systems are typically implemented in situations involving safety-critical use, where a governance model is required that both provides technical resilience and ethical integrity and legal responsibility. This paper focuses on the regulation of high-risk AI in healthcare through the lens of the relationship between system reliability, transparency, and regulatory accountability. It discusses the main aspects of technical robustness, such as accuracy, explainability, bias reduction, and cybersecurity, and determines how these technical norms are in conflict with the ethical foundations of patient autonomy, fairness, and human control. The paper also evaluates current legal and regulatory frameworks to define accountability, liability and compliance loopholes in AI-driven healthcare applications. The research provides a combined governance framework that harmonizes the technical, ethical, and legal approaches to innovations by synthesizing the three aspects of action to attain patient safety, and trust in the population. The results add to the discussion of responsible adoption of AI in healthcare and offer the policy-related implications to the regulators, healthcare facilities, and AI developers aiming to establish trustful and responsible AI systems.