Robots as animals: A framework for liability and responsibility in human-robot interactions

As robots become more common across society, there is a pressing need to deal with questions of moral responsibility and legal liability in accidents involving semi-autonomous and autonomous machines. Previous attempts to address these questions have assumed machines with either minimal autonomy or full intelligence, and thus have not adequately considered the current and likely future state of the art in robotics and artificial intelligence. In this paper, we offer general principles to make sense of the foregoing issues, and propose a framework for addressing questions of responsibility and liability in human-robot interaction. This approach is based on the premise that robots can be analogized to animals for the purpose of assigning responsibility and liability when robots are involved in accidents. We provide justification for this approach, consider its implications, and discuss several of its advantages in analyzing human-robot interactions.

I. Introduction

We live in a time in which robots are increasingly found outside the factory and inside the home or office or even on the battlefield, performing simple but useful tasks. In fact, the state of the art in robotics technology is comprised of systems that have limited, albeit steadily increasing, autonomy. We may reasonably call these systems “semi-autonomous.” As the technology grows more advanced, the ethical questions involving these systems become more pressing. In particular, the emergence of home and office robots raises the question of responsibility for robot accidents: When a robot is involved in an accident, whom (or what) do we blame? The robot, its manufacturer, the owner, the victim – all are likely candidates. However, current ethical thinking is far from settled on the extent to which responsibility should fall with each party in a robot accident, let alone who should be legally liable in such cases.