Humans are not the only persons in the eyes of law. Jurisprudence has evolved over the years to create legal fiction and recognise body corporates and associations of natural humans as capable of possessing personhood and termed them as “persons”, capable of exercising rights and holding them accountable for their actions. The most recent development in this area has been the debate whether AI should be granted personhood in the legal framework. In the last few years, this debate has assumed practical implications with the rise of AI based technology and the increase in creation of robots and AI machines capable of performing human functions.
In 2017, Saudi Arabia made the headlines by granting ‘citizenship’ to Sophia, a humanoid robot. While it was still unclear what it meant to be a “citizen”, it was speculated that Saudi Arabia might create personhood to confer on the robot. In the same year, the European Parliament passed a resolution to consider the creation of a “a specific legal status for robots in the long run, so that at least the most sophisticated autonomous robots could be established as having the status of electronic persons responsible for making good any damage they may cause, and possibly applying electronic personality to cases where robots make autonomous decisions or otherwise interact with third parties independently” An important moment in this debate happened when WIPO on December 13, 2019 published its issues paper, and made a call for comments from the global audience. Following this, WIPO also held virtual conferences on the issues surfacing in the realm of AI, IP and their intersection.
The various theories of personality embedded in legal jurisprudence makes it possible to grant personhood to AI, and the entire corporate personality theory is proof of the same. However, the question regarding the feasibility of vesting rights in AI and also holding it liable for its actions rests on more nuanced issues. When we talk about granting legal personality to AI, essentially, we are talking about creating a body corporate which can act on its own but with a human being behind it. So, should we endow machines with such a personality? While it may help in better formulation of laws regarding liabilities and responsibilities, the necessity to confer personhood in order to grant rights is debatable. Further, to penalise AI, intention or motive will have to be proved and those are concepts uniquely human in nature.
John Chipman discusses the classical notion of legal personhood in the “Nature of Sources of Law”. In his opinion, in common parlance as well as in most books including law books, the term “person” invariably denotes a human. However, considering a technical legal viewpoint, there are rights and responsibilities attached to the term.
The creation of an intelligent robot and its rise to monstrosity has been the premise of several stories in popular culture, be it Frankenstein’s creature, who sought revenge for non-acceptance in society, or Shankar’s Chitti 2.0 who could not handle his “humanly emotions”. AI has the potential to exceed their creators’ imagination and foreseeability. This potent mixture of complications makes lawmakers wary of granting them personhood which could be equated to that of a natural person.
While discussing the concept of agency and personhood vis-a-vis intelligent agents, Schirmer calls it a “dual dilemma”, and explains that while it is hard to ignore the conceptual and practical arguments to grant intelligent agents a status similar to humans; at the same time, it will be difficult to justify withholding certain rights and obligations from the intelligent agents. Schirmer’s arguments stem from Germany granting a partial legal status to intelligent agents.
The need of the hour is to bridge the gap between the legal issues which arise out of the complexities of AI and the solutions that need to be formulated to address them. Surprisingly, no matter how advanced technology gets, there still is scope for development to make it a completely efficient working model, and thus the need for a legal framework. Also, given the fact that the existing legal framework has not evolved as quickly and substantially as compared to the rapid progression in AI technology, it is to be understood that its capacity to function at the optimum level has been capped, and needs to be channelized in a proper manner.