STRICT LIABILITY OR FAULT-BASED REGIMES FOR AI-CAUSED HARM? A DOCTRINAL ANALYSIS ACROSS COMMON LAW AND CIVIL LAW SYSTEMS

Abstract

The rapid deployment of artificial intelligence systems across critical sectors has created unprecedented challenges for traditional tort law frameworks. When autonomous vehicles cause accidents, medical diagnosis algorithms produce errors, or automated decision systems generate discriminatory outcomes, fundamental questions arise about liability attribution. This research examines whether strict liability or fault-based regimes better address AI-caused harm through comparative doctrinal analysis across common law and civil law jurisdictions. The study analyzes existing liability frameworks in the United States, United Kingdom, Germany, France, and Japan, evaluating how each system's principles apply to AI-specific scenarios. Through examination of recent case law, legislative proposals, and doctrinal scholarship, the research identifies critical gaps in current frameworks where AI's autonomous decision-making, opacity, and distributed responsibility challenge traditional causation and fault concepts. Findings reveal that pure fault-based approaches struggle with AI's "black box" problem and difficulty establishing negligence standards, while strict liability faces challenges in defining "defect" for learning systems and determining liable parties in complex AI supply chains. The research proposes a hybrid model incorporating elements from both regimes, with strict liability for high-risk AI applications and modified fault standards for lower-risk contexts. This framework addresses the unique characteristics of AI while maintaining accountability and encouraging responsible innovation across different legal traditions.