Artificial Intelligence, Accountability and Criminal Responsibility: Regulatory Challenges for India's Digital Economy

Abstract

With the increasing use of artificial intelligence in sectors ranging from self-driving cars, health care diagnosis, financial markets, and even lethal autonomous weapons, there is a noticeable inadequacy in the current criminal liability framework. With the traditional doctrines of criminal liability based on actus reus and mens rea, coupled with the principle of individual responsibility, the current legal regime fails to provide answers to the challenges posed by autonomous machines which function with varying degrees of human involvement. The present paper will focus on the intricate issues pertaining to the criminal liability of AI-powered actions in the Indian legal regime. The paper then goes on to differentiate between those instances where AI is merely used as a tool, and responsibility is attributed to human agents like programmers, manufacturers, and users, and those cases where autonomous AI agents come into the picture, rendering attribution difficult owing to the “black box” problem and issues of culpability of intent. A comparison of legal personhood of AI in different jurisdictions is undertaken, with special focus on the proposal for electronic personhood in the European Union. A discussion of the difficulties faced in each domain, namely, autonomous vehicles, health care, white-collar crimes, and content creation, is followed by some suggestions for reforming Indian criminal law, which include the amendment of the Indian Penal Code and the Information Technology Act.