Liability and automation: legal issues in autonomous cars

The deployment of highly automated systems, such as autonomous cars, is going through an accelerated expansion: as usual for emerging disruptive technologies a slow start is followed by a more and more rapid surge. One of the most important legal issues concerning these systems is related to liability for accidents. In particular, highly automated systems will make choices and engage in actions – usually with some level of human supervision, or even without any such supervision. In this context, there is the need to analyse how the decision-making process is split between humans and machines, and critically revise the way tasks, roles, and liabilities are allocated. In this contribution we analyse the impact of automation in the allocation of liability within autonomous cars. We first discuss the tasks allocation between human and automation, and the resulting responsibilities. Then, we analyse how the introduction of different levels automation gives rise to a redistribution of tasks between human and automation and, therefore, a reallocation of the liability burden between the user and the manufacturer.