Research on Tort Liability in the Context of Medical Artificial Intelligence

Abstract

The rapid integration of artificial intelligence (AI) into healthcare has substantially improved diagnostic accuracy and treatment efficiency while simultaneously creating complex challenges for determining tort liability. The autonomous decision-making capacity of medical AI systems and the opacity of algorithmic processes (“black box” effect) complicate the identification of fault, the establishment of causation, and the allocation of liability within traditional tort law frameworks. This article argues that existing legal mechanisms are insufficient to address harms caused by highly autonomous medical technologies and proposes a comprehensive liability regime adapted to the specific risks of medical AI.

The proposed framework is based on three key dimensions. First, specialized legislation should introduce classification and risk-grading standards for medical AI and clear criteria for identifying algorithmic defects, thereby clarifying responsible parties. Second, a coordinated system of imputation combining fault-based, strict (no-fault), and equitable liability is necessary to balance technological innovation with effective protection of patients’ rights. Third, a diversified remedial mechanism should include reversal or mitigation of the burden of proof, compulsory liability insurance, and compensation funds to ensure prompt and adequate compensation.

Such a risk-oriented legal framework is essential for safeguarding patient safety and supporting the sustainable development of medical AI.