Priyam Jhudele, student at National Law Institute University, writes on different models of liability for artificial intelligence entities.

Abstract

Humans have taken giant strides in the field of robotics in recent years. From self-driving cars to interactive software like Siri, terrific progress has been made. However, as we move forward towards smarter machines, machines that can think and react, we need to be wary of the threat they pose to mankind. This is the reason why renowned experts like Stephen Hawking and Elon Musk have expressed their concern over the increasing use of artificial intelligence. Moreover, the cases of robots committing crimes are on the rise, whether those crimes are committed directly or indirectly. We cannot impose liability for a crime committed by an intelligent thinking machine on a human merely because he programmed it. We need to devise a system of liability for punishing the intelligent machines which can think on their own. Thus, this paper proposes three models of liability under which artificially intelligent entities can be punished. It also proposes tests which can be used for determining their intelligence and the appropriate punishments which can be given to such entities. It also suggests changes which can be brought about in the Indian law to bring such entities under the purview of law.