The Applicability of Product Liability to Artificial Intelligence Systems

Abstract

In this article I examine the applicability of the Australian Consumer Law provisions on product liability for defective goods to artificial intelligence (‘AI’) systems. I argue that significant reforms are necessary to effectively address defects in AI systems. Key areas for reform include clarifying the term ‘component’ of goods to encompass digital elements such as AI-related services and revising the classification of manufacturers to reflect the roles of various parties in AI development and deployment. Most critically, the test for assessing defectiveness should be reconstructed to capture the distinctive characteristics of defects in AI systems. Moreover, the scope of actionable damage should be expanded to ensure comprehensive protection for individuals harmed by defective AI systems. Drawing on comparative insights from the European Union and the United States, I suggest that the successes and limitations of these jurisdictions in addressing liability for defective goods provide valuable lessons for Australia to strengthen its product liability regime in the age of AI-driven technologies.