Attributing legal personality to artificial intelligence : a possible solution?

Abstract

This dissertation seeks to determine the possibility of vesting artificially intelligent entities with legal personality and its adequacy as a means of regulating the technology. The first chapter serves to introduce the subject by discussing AI and its rapid development, with particular emphasis on its disruptive nature. This is supplemented by a number of case studies to highlight the need for legislation dealing with AI. Furthermore, a section of the chapter is dedicated to developing a proper definition of AI for the purposes of its legal regulation, which is to be used throughout the dissertation. The second chapter introduces the challenges posed by AI to existing legal principles, with particular emphasis being placed on what is referred to as the “responsibility gap” caused by AI. In carrying out this exercise, a number of possible liability regimes that have been suggested by several authors are explored. The purpose of this exercise is to show the inadequacy of these legal frameworks whilst laying the foundation for the discussion on vesting legal personality to AI in the remaining chapters. The next chapter introduces the concept of legal personality with the aim of discussing whether AI can be vested with such personality. The fourth chapter focuses on whether AI should be vested with legal personality by delving into a number of arguments in favour and against the said approach, whilst the following chapter centres around practical challenges relating to its implementation and any possible solutions to such concerns. Finally, the conclusion serves to tie everything together by stating, in the opinion of the author, whether AI should be vested with legal personality or not based on the merits of the arguments discussed in the previous chapters and on a cost-benefit analysis of implementing such an approach, whilst giving reasons for this decision.