This paper analyzes the link between the legal concept of “person” and the status of artificial intelligence in legal systems from a conceptual perspective. In this framework, some triggering questions include: Is there any space in which an artifact or human construction can be considered analogous to a human being? Is a conceptual category possible that can at some point allow for the linking of human actions and those generated by an artificial intelligence? Is there a common denominator that enables the meeting of robots and men / women? If there is, what (if anything) does the recognition of artificial intelligence say about the common denominator? What kind of social reaction gives rise to it and grounds it?

The starting point of this research is that, in the field of law, the concept of person has the potential to establish itself within a common scenario in which human beings and other kinds of entities act, as long as its meaning is established in such a way that enables a common interaction. To explain this, it is worth thinking that, in the legal field, “person” is often used in a merely technical sense: an entity capable of acquiring rights and obligations; but also, in many other definitions, this concept is used in a common way and linked to a moral prescription: the human being who deserves special legal treatment for having such a nature. In this framework, here it is described how, depending on the meaning adopted regarding the legal concept of person, artificial intelligence may or may not be conceived as a subtype of it. Likewise, some characteristics and consequences that the incorporation of the “electronic person” has on the legal concept of person are explained.

To do so, a comparison is made of the regulation proposed by the European Parliament with three types of persons recognized in the Argentine legal system: human persons, non-human animal persons, and legal persons.

The main debate, from a philosophical point of view, lies between the positivist and non-positivist (in a broad sense) positions. It is explained that depending on the perspective regarding what it is to be a person in law, problems may arise in admitting the incorporation of electronic persons under the common denominator. A positivist position will have no inconvenient: as long as the legal system attributes rights or obligations, there will be a person. Thus, we can have people with all the rights and obligations (such as human persons), people with some rights and obligations (such as legal persons), people with some rights (such as non-human animal people), and people with some obligations (electronic people). There are not too many characteristics in common between all of these parties. Perhaps the most conspicuous commonality between these groups is being the recipients of rights and/or obligations. We can also mention some kind of ability to interact with the environment. 

Non-positivist positions are more varied. Some positions might recognize the usefulness of incorporating electronic persons. However, they could object to this admission due to the effect it has on the legal concept of person in general.

As a conclusion, it is stated that the notion of electronic person can cause theoretical damage to the semantics of the legal concept of person. On the one hand, this is because the concept may lose part of its prescriptive force at the same time that it reduces the definition to the current casuistry which is conditioned to the emergent. On the other hand, it has a practical advantage, which is to provide a tool to attribute liability for damages to an entity that is interacting with companies. The question is thus, as in the case of human persons, legal entities and non-human animal persons: is the concept of person the most useful and most convenient to achieve the desired effect with respect to artificial intelligence?