Employer’s Vicarious Liability for Damage Caused by an AI Worker: Comparative Law Perspective

The article examines whether the proposal to apply vicarious liability is the optimal method of regulating the employer’s liability towards third parties for damage caused by AI systems. The research concerns autonomous AI systems used by the employer to achieve the same goals as previously could be achieved only by the human workforce. In this article, the concept of the optimal solution refers to a liability model capable of addressing the challenges posed by the inherent characteristics of AI, such as autonomy, complexity, opacity, the ability to learn, and interaction with the environment, including humans. This model should ensure that injured parties have a genuine opportunity to receive compensation. Scholars have proposed the use of vicarious liability as, in their opinion, the optimal method to regulate employers’ liability vis-à-vis third parties for damage caused by autonomous AI systems. Experts predict that applying such systems on a large scale is only a matter of time. This article concentrates on the civil liability of an employer in situations where damage was sustained by a third party who is not in any legal relationship with the employer, and the damage was not caused by a defect of the AI system. The analysis covers the legal perspective of the EU and the US.1