Recognising Autonomous Systems as Electronic Persons – Part 1

Adyasha SahooLaw

Recognising Autonomous Systems as Electronic Persons

Amid our digital era, where silicon and soul converge, and algorithms permeate our daily lives with an uncanny resemblance of intelligence, a compelling question arises: Can we confer personhood upon our silicon companions? Imagine a world where you arrange a coffee date with a robot friend. Imagine a courtroom where a robot stands trial, its circuits pulsing with anticipation, or a scenario where your silicon partner breaks your heart.

Electronic personhood is a captivating realm where the concept transcends mere legal jargon. The journey comments at the intersection of jurisprudence and circuitry. As we explore the historical definitions of person and legal person, we come across the birth of electronic jurisprudence. This new frontier presents us with intriguing challenges. Can an algorithm experience remorse? What happens if Siri demands a lawyer? This two-part article seeks to unravel the electronic personhood, the legal corridors where the silicon meets the soul.

Definition of “person”

From the jurisprudential perspective, there are four possible logical deductions based on the work of various scholars.

  • First, a person is a being, i.e., the factual existence of the being qualifies him to be called a person.
  • Second, the being should possess rationality and an advanced level of brain functioning. They should clearly distinguish between right and wrong.
  • Third, it is possible that some persons or significant groups of human beings do not qualify as persons by birth since they may not have an advanced level of functioning.
  • Fourth, certain non-humans can exist, but they will be considered legal persons if they meet the criteria of being persons.

Before we proceed further, some considerations deserve mention. In the context of the third logical deduction, the legal system would treat such persons differently. They may not have rights, duties, or remedies. For example, newborns are natural persons but not legal persons. Their consideration as one is for limited purposes only. For example, Section 20 of the Hindu Succession Act, 1956, governs the rights of an unborn child. This provision says the born child and the foetus in a mother’s womb shall have equal status and the same right to inherit property of the deceased.

Similarly, Section 30 of the Transfer of the Property Act, 1882, allows a person to transfer their property to an unborn child. Even though this provision does not consider a foetus a living person, it permits the transfer. Some other examples of non-persons include adults suffering from dementia or individuals with severe brain injury.

This raises a valid question. Can autonomous systems be considered legal persons based on the fourth logical education? This debate, however, does not end here. Those who meet the criteria of being a person get the advantage of possessing moral rights and privileges. They can exercise their own choices and autonomy as much as possible. So, to what extent will it be correct from the legal perspective to grant autonomous systems with legal personhood?

Legal Jurisprudence on this Issue

Gaius, a prominent Roman jurist, was the first to differentiate between person and things. He only considered those humans as persons who can perform actions and be held accountable. Cicero, a Roman statesman and lawyer, acknowledged Gaius’ thinking that only human beings can be considered persons. Cicero further listed four attributes: rationality, personal attributes, status, and profession. Any human being with these attributes is suitable for their consideration as a person. Persons are authors of their actions, and those who are not, their guardians can represent them. Parents representing minors is a classic example of this.

The orthodox view equates a person’s legal personhood with his holding of legal rights and bearing of legal duties. This view worked in the nineteenth century. Still, with changing times, jurists and researchers are trying to make the definition of legal personhood more inclusive, including the emerging autonomous systems.

From the modern learning perspective, any being can be a suitable right holder and a duty bearer on whom an authority can impose liability. The Black’s Law Dictionary defines a legal person as a legal entity given certain legal rights and duties. A being real or imaginary who, for legal reasoning, is treated more or less as a human being. The fundamental division of legal personhood that has existed since the nineteenth century is natural, artificial, and juristic persons. If a being qualifies as a natural person, it is an individual who is a legal person.

Visa AJ Kurki, a legal scholar and philosopher, introduced the bundle theory of legal personhood in his book. His theory challenges the orthodox view of legal personhood. He argues that legal personhood is not based on particular sticks (rights) alone. Instead, he divides the cluster into various active and passive incidents, forming the concept of personhood. Active incidents are the activities a legal person can do actively, such as signing contracts and exercising their voting rights. Passive incidents are activities that happen to a legal person, such as the law protecting them against a wrong.

Kurky’s theory challenges the distinction between legal persons and non-legal persons. It presents a loophole when considering entities like animals, trees, or autonomous systems that may not fit the orthodox mould of personhood but still have some rights or incidents associated with personhood. To find the answer to this, we need to identify an acceptable definition of autonomous systems and whether they qualify for the characteristics of legal personhood.

Recognition of Autonomous Systems as Legal Persons

In philosophical terms, autonomy refers to responsibilities capable of being attributed only to a moral agent. In today’s time, morality shapes a person’s subjective freedom to decide what is rational for them. Emmanuel Kant, the philosopher, defined morality as autonomy by breaking the paradigm of morality as obedience. When a person, through this moral quality, rejects the inequality within, he is capable of judging his actions without any external intervention.

However, if one understands autonomy in a strong sense, no robots or AI qualify for these conditions. In another sense, autonomy can be understood as the ability to act without human intervention. Most autonomous systems that exist today need human intervention in some form. This is autonomy in the weaker sense. For example, this form of autonomy exists in autonomous drones, driverless vehicles, and robotic vacuum cleaners. This is an appearance of agency.

As per the classic definition by Richards & Smart (2013), robotics artefacts are analysed from a weaker sense of agency, not to be confused with its strong sense. In this aspect, a robot is a built system that displays only physical and mental agency and is non-existent in the biological sense. Robotic artefacts only have material existence, make rational decisions (weak or apparent autonomy), and are machines.

In the verbalised legal world, “legal person” is an autonomous centre of legal relations. In 2017, the European Parliament proposed creating electronic personhood for intelligent robotic artefacts. The resolution established that a robot should be considered intelligent when it has the following characteristics:

  • Existence of sensors allows it to exchange data with the environment
  • Ability to learn from experience and interact with the environment
  • Existence of material support
  • Ability to adapt
  • Absence of life in the biological sense

Conferring legal personhood on companies is a model justifying legal personality recognition for autonomous systems. For example, Turner (2018) argues that possible abuses, such as the lack of responsibility of programmers and engineers, can be disregarded by piercing the corporate veil. However, this argument also points out the problems present in the corporate personality model.

We will continue this discussion in the next part of this article and elaborate further on potential challenges in recognising personhood for autonomous systems.