Who Is Liable When AI Misdiagnoses? Reconstructing the Legal Accountability of Physicians and AI Developers in Indonesian Healthcare

Abstract

The increasing use of artificial intelligence (AI) in clinical decision-making presents significant challenges for healthcare liability law. AI-assisted diagnosis may improve efficiency and diagnostic accuracy, but errors generated by algorithmic systems raise a fundamental legal question: who should be held accountable when an AI-supported diagnosis causes patient harm? This article examines the allocation of legal responsibility between physicians, healthcare institutions, and AI developers when AI contributes to diagnostic errors in Indonesia. Using a normative juridical approach, the study analyzes Indonesian health law, medical liability principles, patient protection frameworks, and emerging approaches to AI governance. The analysis demonstrates that Indonesia's existing legal framework remains predominantly human-centered and does not clearly regulate liability arising from autonomous or semi-autonomous clinical technologies. Physicians may remain responsible for clinical decisions, yet imposing liability exclusively on physicians may be problematic where errors originate from defective algorithms, inadequate training data, or undisclosed system limitations. This article proposes a distributed liability framework based on the degree of human control, foreseeability of harm, system transparency, and the respective duties of physicians, healthcare providers, and AI developers. Such a framework is necessary to ensure patient protection without discouraging responsible technological innovation in healthcare.