By Pratyush Pandey, a IVth year B.A. LLB (Hons.) student at Rajiv Gandhi National University of Law, Patiala
Introduction
In recent years, a call for granting personhood to Artificially Intelligent Systems (hereinafter, A.I.) has gained a sudden, yet great, impetus. Nations like the United States and the European Union have already started granting juridical personality to A.I., employed in works of constructive developments. In the contemporary world, there exists a long list of “non-human entities”, like Rivers, Forests, Ecosystems, etc., which have already been accorded Legal Personhood. This leads to the question, “Whether A.I., considering its exuberant usage, is eligible to be accrued Legal Personhood?” In light of these prefatory submissions, the article analyses the concept of Legal Personhood, and, thus, answers the question raised above.
Artificial Intelligence
Minsky and McCarthy, whose brainchild A.I. was, in the 1980s defined it as an intangible machine, based upon binary programs, that could perform any task which a human could do with an application of intelligence. Thus, A.I. could be comprehended as a system that could demonstrate behaviors associated with human intelligence and creativity, such as learning, reasoning, planning, and execution. However, this definition of A.I. faces criticism for being highly exclusive in terms of “defining characteristics”, which are restricted to only intelligence and creativity[Ed1] , in order to qualify a system as A.I. Therefore, there is indeed a need for a new definition, suitable and sufficiently broad in the modern parlance.
Legal Personhood
The idea of Personhood has emerged from the concept of Person. Salmond defined Person, as “one capable of holding rights and duties”.[i] Notably, for Salmond both “humans” and “non-humans” were qualified to be people if they are capable to hold rights. Thus, from this progressive definition, emerged the classification of Natural Person and Juridical Person. In light of this classification, any ‘non-human entity’ like corporations, idols, ecosystems, etc., may be accorded the status of “Juridical Person”, accruing it Legal Personhood based upon its recognition or utility in the society.[ii] Therefore, Jurists, all around the globe, have conceded that jurisprudential theories on Person and Personhood do recognise and allow recognition of A.I. as Juridical Person. Moreover, they also aver that under the postulates of Fictional Theory, A.I. systems may indeed be accrued with legal personhood.
Global Trends on Legal Personhood to Artificial Intelligent Systems
As stated, the United States and the European Union are forerunners in accruing Legal Personhood to A.I. The progress made by the US and the EU, in the specified regime, is evident from their academic discourse, research, and legislative approach. Hereunder, we would discuss the steps taken, so far, in both these jurisdictions.
To start with, Europe, in 2017, put forth and passed the Civil Law Rules on Robotics Resolution to grant A.I. the status of ‘Electronic Person’. Provision 59-F of the said Resolution posited that sophisticated autonomous A.I., employed in various industries, possesses the status of electronic persons, where they make autonomous, independent decisions, and, therefore, be made responsible for any loss or damage they make against a third party. In tandem, the Resolution also brought A.I. under the regulation of Civil and Criminal Law, for rights and liabilities. Moreover, to facilitate the Resolution the EU Commission on Artificial Intelligence, in 2018, released a set of guidelines to concretize the idea of Legal Personhood to A.I. by way of the appropriate, ethical and legal framework.
Further, in the United States, the Manual of Law of War of the United States DoD has also recognized A.I. as Juridical Person. Herein, under provision 6.5.9.3, certain regulatory obligations are laid down (though, not mandatory), with regards to the conduct of autonomous war-machine during the war. The Manual avers that when a war machine is capable and independent of making ‘factual determinations’, and no human agents are attached to it, then, in such cases, the machine is accrued with legal personhood, and is solely responsible for any of its actions.
Notably, in both these jurisdictions there exists stark similarities in the approach of granting Legal Personhood to A.I. For instance, both these nations follow codified regulations for granting legal personhood to A.I. Moreover, both follow similar notions of jurisprudence with regards to Juridical Person and Legal Personhood, as discussed above. Moreover, in both these jurisdictions, A.I. may either be accorded Legal Personhood or be given the status of an “agent of business”, bringing it within the ambit of Civil or Criminal liability.
In totality, both the US and the EU have proved themselves to be progressive enough to evolve with the ever-modernizing technological advancements, and have laid down a proper precedent for other nations to follow upon.
Granting Legal Personhood to Artificially Intelligent Systems – The Indian Way
With regards to road–map, a plethora of nations, like New Zealand[Ed2], Australia, Canada, etc., have already laid down policies for granting Personhood, allowing ‘non-human entities’, like rivers, forests, ecosystems, animals, etc., to be accrued with Legal Personhood, and, therefore, the same may be taken as precedents. Therefore, firstly, the Indian policymakers should delve into an exhaustive study of the global practices on granting legal personhood, and then develop an Indian approach. It is only after developing an Indian approach, that the discourse on granting legal personhood A.I. would be tenable and fruitful.
Therefore, the Policy Makers should first develop an Indian jurisprudence on granting Personhood to non-human entities, in line with global practices. Thereafter, they should analyze the worldwide legal innovation pertaining to A.I. laws, like in the US and the EU, and try to develop an appropriate Indian legal framework. Moreover, emphasis should be placed upon the notions of fundamental rights and human rights, so as to maintain a balance and status quo between the rights of Natural Persons and Juridical Persons.
Conclusion
As evident, presently there exists no definite regulations or guidelines for determining the Legal Personhood of A.I. The A.I. Laws, even in the most developed nations, are at a fledgling state and have a long evolution to go through. Therefore, it is important for the developing nations to come up with their jurisprudence on A.I., to set precedent and help develop a global jurisprudence. For India, it is a matter of utmost urgency to acknowledge the gravity of the jurisprudence it is lacking in, and start developing an Indian Jurisprudence on Personhood. It is only after this development that India could become a part of the global discourse on A.I. Laws. For every nation, it is also important to keep track of the effect of A.I. and A.I. Laws on its society, its laws, and its procedural legal framework, for it is the most important aspect of making a conducive, comprehensive, and compatible law. The discourse of Legal Personhood of A.I., thereby bringing machines-of-intellect within the ambit of legal liability, is a discourse for the future, and, therefore, it is most important that the laws we make are future-ready and flexible enough to avoid any ‘judgment-day’.
[i] John W. Salmond, Jurisprudence or The Theory of Law355 (4th edn, 1913).
[ii] P.J. Fitzgerald, Salmond on Jurisprudence 308 (12th edn., 1966, Indian Reprint).
[iii] Ram Jankijee Deity v. State of Bihar, (1999) 5 SCC 50.
[iv] Salim vs State of Uttarakhand, Writ Petition (PIL) No.126 of 2014 (Decision stayed by Supreme Court in July, 2017)