As robotics and artificial intelligence (AI) become more prevalent in the world economy and modern society, timely legislative efforts could help smoothen the proliferation of this technology and prevent undesired consequences.
From relatively humble beginnings limited to mere remote-controlled automation, AI has developed immensely over the past decades and the prospect of fully-autonomous human-like artificial intelligence is becoming a day-by-day reality.
Following large efforts in the blockchain sphere and other moves to become a trailblazer in the digital economy, the Government of Malta has recently announced that it would be exploring the possibility of granting citizenship to robots.
As to legislative efforts in this regard, this is by no means unprecedented. In 2017, the European Parliament’s Justice and Human Rights committee (JURI) discussed an own-initiative report by MEP Mady Delvaux, which inter alia toyed with the idea of granting robots an “electronic personality”.
Brushed off by many as an attempt for creators and developers of such robots to wash their hands clean from liability and responsibility over the actions and consequences of their creations, this discussion has yet to lead to any tangible legislative action.
Granting some form of legal personality to robots and AI could have very serious implications of both a legal and ethical nature. Should robots be granted legal personality? And if so, what type of legal personality should they be granted?
It is not unheard of that non-human entities are granted legal personality. The idea that an organisation could have a legal personality of its own, distinct from that of the natural persons behind it, has been in vogue since ancient Rome.
Pope Innocent IV is considered to have popularised the idea of the persona ficta (or fictitious person) as a means of reconciling the infrastructural needs of monasteries with the vows of poverty taken by their monks. Thus, through the legal construct of a fictitious juridical person, monasteries were granted certain rights and obligations which had been hitherto largely restricted to natural persons. This idea laid the foundations for the legal personality of modern corporations.
The concept of a “juridical person” that is not also a natural person, yet enjoys certain legal rights and duties, which are otherwise restricted to natural persons, is the cornerstone of the modern economy.
Corporations do not enjoy the totality of rights that natural persons enjoy; they cannot marry or vote, but can enter into contractual relations, generate wealth, sue, hire staff and open bank accounts, but must also pay taxes and can be held liable for civil wrongs, regulatory breaches and in some instances, even crimes.
However, ultimately, in real terms, corporations can by no means be considered to be entirely distinct from the natural persons behind them. They are not rational beings and are completely unable to act, were it not for their human agents. When the law or authorities require a corporation to act or refrain from acting in a certain way, it is essentially its human agents that must follow through.
Sanctions levied against a corporation, are ultimately suffered by its human agents. Moreover, the separate legal personality is not absolute and if the context so requires the corporate veil may be pierced to reveal the natural persons who must answer for the corporation’s action or inaction.
While AI and robots are becoming increasingly autonomous, ultimately they have come into existence through the endeavours of natural persons, and many a time require some form of input from natural persons in order to operate.
Just as corporations are ultimately an expression of the natural persons behind them, the same can be said for AI. We must thus tread with caution before rushing to grant legal personality to AI, as the implications could be immense. Ultimately, AI lacks the “will” (volontà) to act completely independently, as it can ultimately always be traced back to a natural person who programs or controls it.
The legal and ethical implications of granting legal personality to AI are immense, and while we have not reached a technological stage which requires the implementation of Asimov’s laws on Robotics, there are a vast array of pitfalls to look out for particularly in relation to product safety, data protection, anti-money laundering and civil and criminal liability.
The potential benefits of harnessing this sector responsibly are enormous, yet we must act with caution when legislating. Thus far, one has yet to understand what benefit would result from granting AI a separate legal personality, yet if such does indeed turn out to be favourable, it is of utmost importance that a causal link to natural persons be established, much in the same way that this exists in relation to corporations.