Analysis of the Legal Subject Status of Artificial Intelligence

The large-scale application of artificial intelligence in social life has brought about new legal issues such as the attribution of rights and the definition of responsibilities. It is necessary to discuss whether artificial intelligence can be used as a legal subject. In view of the existing disputes on this issue, we need to re-examine it from the perspective of second-order observation. Human beings are the origin of the mechanism of the subject, possessing rationality and free will, and providing the source of value for the subject of law. Human beings are the original subject of law. In the process of substantiation, the legal subject is separated from the personality characteristics, and the legal person as the derivative legal subject appears. Artificial intelligence does not have the condition to become an original legal subject, but if it can meet the long-term fundamental interests of human beings as a legal subject, there is still the possibility of being formulated as a derivative legal subject.


Introduction
In recent years, the development of artificial intelligence has set off a new trend.In 2016 and 2017, the AlphaGo, developed by DeepMind successively defeated Chinese and South Korean Go masters in the man-machine contest, showing the strength of artificial intelligence to the world.Artificial intelligence has been extensively and deeply applied to society and our life, gradually changing the operational scene and model of social relations and becoming the driving force of industrial and social revolution.For example, artificial intelligence in the financial field called intelligent investment advisers which can sign a contract and

Different Views on Artificial Intelligence as a Legal Subject
At present, most of the researches on the legal subject status of artificial intelligence focus on exploring the concept of artificial intelligence from different angles, and then draw a conclusion that artificial intelligence is a subject, an object or a compromise between the two.Opinions on the legal subject status of artificial intelligence generally fall into the following categories: First, negative theory.
The theory holds that artificial intelligence is only the object of legal relations and should not be granted legal subject status.As the development of artificial intelligence has not posed a subversive challenge to the traditional subject of law theory, we should still adhere to the traditional theory in the short term, and should not define it as the subject of law.The second is positive theory.The theory holds that artificial intelligence should have the qualification of legal subject, including the agency theory, fictitious personality theory, electronic personality theory and other specific types.Scholars who hold the positive theory believe that with the rapid development of artificial intelligence, it will be widely used in all fields of society and have the ability to independently affect the rights and obligations of others.With the development of society, the subject has broken through the limitation of natural persons, and many non-natural entities have been gradually endowed with the qualification of "human" in law based on the needs of lawmakers, and have obtained the subject status.The trend of "non-human seen as human legally" is increasingly strengthened, which therefore explains the existence of humanoid robots to obtain the qualification of legal subject.Artificial intelligence is similar to legal persons.Since legal persons can have legal personalities, artificial intelligence should also have legal personality (Davies, 2011).It is not only the inherent law of the development of rights, but also the inevitable trend of the development of society to recognize and grant robots the subject status of rights.Granting the non-human author/inventor the legal personality will provide a new incentive for human to utilize the creativity of animals and artificial intelligence.Third, compromise theory.It is also known as the limited legal personality theory of artificial intelligence.This theory holds that artificial intelligence has legal personality, but its personality is special, and the scope of its rights and obligations is limited compared with other legal subjects (Yuan, 2020).The "limited legal personality theory" believes that the essence of artificial intelligence is a tool, and its attribute serving the development of human society cannot be changed.However, artificial intelligence has independent and autonomous behavioral abilities, which should be endowed with legal personality.Since the consequences of artificial intelligence's behavioral ability to bear are limited, it shall be deemed to have limited legal personality, and be applied to special laws.On the whole, the compromise theory is not really a compromise, but rather it creates a new way to challenge the dichotomies of "subject-object" and "human-object".It clearly advocates that artificial intelligence has legal personality.Even if it has only partial and limited rights and obligations compared with natural persons, or its own legal subject status is incomplete, or the actual ability to enjoy rights and fulfill obligations and normative qualifications have defects, the premise is to admit its positiveness in qualifications, that is, positive attitude in the subject status, which is fundamental.As for the number of rights and obligations, it is a "quantitative" distinction, and does not affect the "qualitative" judgment, so this compromise theory still belongs to the category of the positive theory.

Analysis of Reasons for the Debate
For legal researchers, whether artificial intelligence can become the subject of law has a wide range of disagreements.Researchers hold different perspectives at the beginning of the discussion, which makes it difficult to reach a consensus.
The main reason lies in the limitation of research perspective.Currently, most of the researches on the legal subject status of artificial intelligence have observed artificial intelligence from the perspective of the legal subject (namely the legal researcher).According to Luhmann's constructivist epistemology, the observation belongs to the "first-level observation", of which the premise is the existence of distinction.The distinction between legal subject and non-legal subject is the premise of our observation.Once we make this distinction, the "legal subject" becomes the marked side, while artificial intelligence, as the traditional "non-legal subject", becomes the marked side.The distinction means that man, as a certain legal subject, regards himself as the only self-referential system and an independent and privileged observer.This kind of observation based on distinction has an insuperable limitation of perspective-the observation cannot observe itself.
As a result, the research on the legal subject of artificial intelligence itself has been neglected for a long time.In view of the limitations of studying the issue from the perspective of first-level observation, "second-level observation" shall be adopted to observe the subject of law and artificial intelligence from the perspective outside the original system, on the basis of the observation of the subject of law, in order to review the relationship between the subject of law and artificial intelligence.

Can Artificial Intelligence Become an Original Legal Subject?
When discussing whether artificial intelligence can become a subject of law, one of the first questions to be answered is what is a subject of law.This requires us to re-examine the subject of law that has been ignored for a long time from the perspective of second-level observation.Although the term "subject" has existed for a long time, the concept of legal subject has not been widely used until modern times, so it has distinct characteristics of modern positivism jurisprudence.
For example, Duguit believes that "the meaning of legal subject generally refers to those who can be the holders of subjective rights, and only individuals with conscious will in the world" are legal subjects.Gray believes that "the technology-based legal meaning of 'subject' is the undertaker of legal rights and obligations".Kelson defines a legal subject as "the subject of a legal obligation or right."However, this does not mean that the subject of law is a purely empirical or technical concept unrelated to value.It is noticed that though the concept of the legal subject is mostly described by the method of positivism jurisprudence, the concept of the legal subject is actually elaborated around "person", whether it is the holder of subjective rights or the undertaker of rights and obligations.This is because the concept of the legal subject originates from the concept of the subject (person) in philosophy.Therefore, we can regard man as an original legal subject.

The Matrix Characteristics of the Original Legal Subject
As the original legal subject, human is the origin of the mechanism of the subject.The word subject comes from the Latin word "subiectum", which originated in ancient Greek and primarily meant base and matrix.The symbol of the real establishment of the concept of subject is the thinking paradigm created by Descartes based on the dichotomy of "subject-object", which started the turning of philosophical research to the philosophy of subject.Descartes took "cogito" and intervened by the subject.In Heidegger's words, the world becomes a "global picture", and human becomes the center of everything.Things other than man become the target represented, recognized and intervened by man as the subject.
The establishment of man's subject status in the philosophical sense opens the way as the subject of law.In the long history of human beings, the origin of thinking and elaborating law did not start with human beings.Law was regarded as having a source beyond human beings.For example, medieval jurists believed that human beings knew the law through the revelation of divine will.In philosophy, the establishment of the subject status of human means that law is truly regarded as a human order.Law is no longer the revelation of God or the gift of nature, but the invention of human beings themselves, and an order of human behavior.It can be said that modern subjective philosophy has clarified a doctrinal value of human beings and has an obvious effect of clarifying the worldview.
Human being is an inseparable existence, not attached to other things and clearly different from other things, has higher value than other things, and is a selfmastery of sovereignty and discourse power of existence."In the eyes of the law, we regard man as the subject, rational and the holder of divine, non-derogable rights."(Supiot, 2019) In the field of law, people determine the basis point and demarcate the boundary.At this time, as the subject, people are not the vassal of God or the co-owner of things, but the self-consistent body of values.They are also the source, the center of law, the real creator, participant, implementer and follower of law.The metaphysical part of law only comes from the value of man himself, and the objective practical part of law only comes from the social life of man.

The Rational and Volitional Characteristics of the Original Legal Subject
Although it is of great significance to establish human as the matrix of legal subject, it is only the first step in the development of subject theory, which means that the existence of subject only as matrix is not enough.The construction of self by subject has just started and is far from complete.Descartes created the dichotomous structure of "subject-object".After dividing the boundary of the two and determining the basic characteristics of the subject, the subject should further improve itself and construct the relationship with the object.In this process, reason and will become the outstanding characteristics of man as the subject of law.Kant regards rationality as human's ability to recognize and understand the objective laws and nature of the natural world, as well as human's ability to identify moral norms and legal rules and guide their behavior.According to Kant's philosophy, human has dual attributes: on the one hand, as an animal with flesh, blood and desire, human is a kind of natural existence, with natural attributes, subject to natural laws, and in a non-free state of domination; on the other hand, as a creature with the ability of higher thinking, human is a rational being, able to understand and transform the world, so as to get rid of the bondage of natural laws, but also put themselves under the norms of rational laws, so as to become a free and self-disciplined ethical subject, and human itself is the end.Human rationality requires the establishment of a field of public and socially implemented legitimate truth, while law is a widely recognized common framework that responds to the common needs of society, interprets the common meaning of human behavior, and integrates subjective rights into objective law.This also makes man the subject of law in a double sense.As the subject of law, human is not only the legislator to establish the law, but also the undertaker of the moral law subject to their own law.As a result of rational guidance, human can make their own choices according to different situations and their reason, and therefore can become the legal subject as the bearer of legal rights.
The subject of law is not only a rational subject, but also a subject with its own will.Puchta said that "man is the subject of law because he deserves the possibility of self-determination, and because he has a certain will."(Kelsen 2021) Reason is closely related to will, which is the presentation of reason on another level, "will … insofar as it can voluntarily choose to act, it is practical reason itself."(Kant, 1991) Some scholars regard will as the "basis, core and soul of the subject."(Li, 2009) If reason lays the foundation of autonomy for the subject, then will is the hub of the transformation from thought to action."There is no will without thinking, will is the practice of thinking, and thinking is the substance of will."(Hegel, 2006) Only with this ability of will, that is, the ability of action, can we express the inner thought embodied in reason through external action.The freedom of will enables people to get rid of the bondage of perceptual desire, have the ability of free choice, and become a real independent subject of autonomy.
Especially when reason builds the fulcrum connecting subject and object, will becomes the channel connecting them.The subject points to the external world as the object through will, and will, as an absolute choice, is a kind of indivisible sovereignty of self.And for this will, as an absolute choice, we usually call it free will.It is the free will that makes us the independent choice of the subject in the law, so that we can become the bearer of legal obligations and responsibilities and become the subject of law.

Whether Artificial Intelligence Can Become an Original Legal Subject
Why does artificial intelligence spark debate about the subject of law?The con- or it can just refer to a set of software systems simulating human thinking.It can be seen that because artificial intelligence is an "artificial product", some researchers regard it as an object from the traditional framework of "humanthing" cognition, but others focus on the "humanoid" or "near-human" characteristics of artificial intelligence, especially from its independent thinking characteristics, and believe that it has the possibility to become the subject.From the above analysis of the subject, it can be seen that artificial intelligence does not have the conditions to become the same legal subject as human.The creation and development of the concept from philosophical subject to legal subject are based on human, which is not only the evolution of metaphysics, but also the historical and practical reality process, reflecting the unique position of human exclusive in the world.Although artificial intelligence has deeply and profoundly affected human social life, it still cannot provide sufficient reasons to become an equal legal subject as human beings.At the same time, due to the limitation of human self-cognition, especially the limited understanding of human brain mechanism, rationality and free will are more reflected in philosophical thinking, which cannot be scientifically described.Therefore, it cannot be proved that the algorithm of artificial intelligence can be completely equal to human rationality, nor can it be proved that the autonomy of artificial intelligence independent of people in certain situations and states is an expression of free will.Therefore, artificial intelligence cannot become the same original legal subject as human beings.

Can Artificial Intelligence Become a Derivative Legal Subject?
Although artificial intelligence cannot become an original legal subject, it is still possible to be a legal subject.This issue is analyzed from the concept of legal subject, especially the two concepts of legal personality and right capacity which are interlinked with legal subject.The starting point of the legal subject is man as the philosophical subject.The legal subject is not a pure concept of legal philosophy, but is gradually formed in the process of reduction and rationalism construction from the philosophical subject of man in the development of positivism.

The Empirical Construction of Derivative Legal Subject
The first step of substantiation of the legal subject is to transform the concrete person into the abstract, and the concept of legal personality comes into being accordingly.The legal subject, starting from individuals, deals with the legal re- on this purpose, the abstract legal personality becomes the greatest common divisor of all people in the law, so as to make the law acquire authority, realize equality and produce wide effects.It is the need of the subject of law to adapt to the development of law as a kind of public rule and universal law.Knipper summarized this deeply abstract process into three steps: first is the purification of freedom normatively by purifying emotion, and the purification of will normatively by purifying will; second is to think about the imaginary man as moral man, or the transcendental vulgarization; third is to achieve unity between rational (legally abstract) man and the mechanized world (Knieper, 2003).The concept of legal personality is the reduction and preliminary abstraction of human beings from the concept of human beings.This kind of abstraction means that the legal subject is transferred from limited abstraction of legal personality to deep abstraction of rights and capabilities, and it is separated from the personality characteristics and the concrete person is abstracted into a rational person, so that the legal subject can be applied to both individuals and groups, and realizes the positive separation of the legal subject from the concept of person.

Legal Person as a Fictional Derivative Legal Subject
This kind of positive divestiture is embodied in reality as the appearance of legal person as a derivative legal subject, and the generation method of legal person as a legal subject is realized by legal fiction.In the opinion of Maine, legal fiction is well-intentioned, a flexible mechanism of legal reality, and an adaptive mechanism of legal development.It is an interpretation method of law, which achieves a balance between legal and social needs.When the civil society and commodity economy develop to a certain stage, there is a practical need to simplify the legal relationship, then the legal person, as a group of people, is fictitious as the legal subject through legal and technical means, which is an inevitable choice of the rational and utilitarian nature of the subject.At this time, rationality shows an instrumental aspect, and the expansion of subject makes it no longer the end itself.They are tools and means to an end.However, when people as subjects start from their own needs to realize the innovation of legal subjects through legal technology, will becomes the self-centered ideological will, and the legal person becomes the legal subject.It is a way for people to rebuild the world image according to their own imagination and expand the absolute self from individual to group.In this process, based on the original subject which is based on human, the legal subjects further develop derivative subjects in the form of legal persons through legal fictions.Savini believed that legal person "is a legal subject confirmed through pure fiction.We call this subject a legal person, a person who exists purely for legal purposes.We find that legal persons are the undertakers of legal relations as well as individuals."Kelson further summarized the deep abstraction process of the concept of legal subject and the birth of legal person as a derivative subject after abstraction.He pointed out that natural person is not a legal person.Not only legal person, but also natural person as legal subject is the result of legal fiction.However, the difference is that natural person becomes a legal person with natural ethical basis, and its being a legal subject has automaticity.As long as it is a natural person, the law will unconditionally recognize its legal personality and confirm its legal subject status.Legal person does not have the ethics of natural person's natural talent, so it lacks natural person's unique natural rights.Therefore, the fiction of legal person needs conditions, and needs to be realized through legal procedures.Therefore, from the point of view of the Pure Theory of Law, if a natural person also needs to be fictionalized to become a legal person and have legal personality, then the fiction of a legal person is double fiction on this basis.The first is to examine whether a legal person meets the conditions of fiction, and it is regarded as a natural person if it meets the conditions of a specific value.The second is to give legal personality to legal persons like natural persons and make them legal subjects (Gray, 2012).

The Possibility of Artificial Intelligence Becoming a Derivative Legal Subject
The most important information that legal person, as a recognized legal subject, transmits to artificial intelligence is that modern legal subject may not be presented as an individual natural person, which means that legal subject can still break through the rigid framework of natural person template and have elastic space in modern times, which provides the possibility for artificial intelligence to become a legal subject.Yuval Harari put forward in A Brief History of Tomorrow that since legal persons can have the subject status, artificial intelligence will inevitably obtain corresponding legal qualifications.Some domestic scholars, inspired by the resolution of the European Parliament on "electronic man" and referring to the approval registration system in the current legal person system design, designed a set of legal procedures such as identity registration, property establishment, legal supervision and termination liquidation for artificial intelligence to participate in legal activities as a legal subject with the identity of "electronic man" (Guo, 2019).The view that artificial intelligence is a legal subject with the identity of "electronic man" is still a minority view, and its rationality needs further scientific proof.However, the differences between "legal person" and "electronic man" can be compared here, so as to better understand the possibility of artificial intelligence becoming a derivative legal subject.The differences between "legal person" and "electronic person" are mainly in the following aspects.First, though both of them have the realistic motivation to simplify the legal relationship and innovate the legal subject when the society has reached a certain stage, the existence of legal person as a natural person group has a much longer history.The basic need to form a group emerged from the initial stage of the development of human society, from the ancient Roman society and syndicate to the religious group in the Middle Ages.Before the legal person system was officially determined in the German Civil Code, the legal practice of natural person participating in legal activities in the form of a group always existed, which is also the historical foundation of legal person being fictitious as legal subject.As for "electronic people", in recent years, as the artificial intelligence has been in rapid development and is increasingly integrating itself into the social life, it has caused specific difficulties in the empowerment or liability attribution under the former laws, which has generated the call to make artificial intelligence as the legal subject.The historical basis of its legal practice is still relatively weak compared with that of legal person.Second, though both of them reflect the external expansion of human legal subject, legal person is still the natural extension of the group form of natural person in society.Based on this, it is much less difficult for the legal person being fictitious as a derivative legal subject.Although legal person is not a natural person in individual form, the individual members of the group are all real natural persons.And its legal acts are mainly realized through the actions of natural persons as legal bodies.But the fiction of electronic man is not so "natural", but a legal subject innovation construction which obviously spans larger and goes further.Electronic man is more like an extension of a certain part of human individuals and its union with the real natural person.In the face of whether this new union form can become a derivative legal subject, human will inevitably treat it with a more cautious attitude.Third, the fiction of legal person is mainly carried out in the field of private law, which mainly solves the legal problems in the aspect of economic relations, and the call for artificial intelligence as electronic person is not only due to the needs of economic relations such as intellectual property rights, but also due to an important reason to solve the problem of legal liability in the aspect of personal infringement caused by artificial intelligence.The latter may involve not only private law, but also the legal issues of public law, which also makes artificial intelligence in the form of "electronic man" the legal subject faced with more complex institutional design problems.In spite of this, if artificial intelligence is to become a legal subject, it can only be realized through legal fiction technology from the legal technology level.The appearance of legal person as a derivative legal subject retains the possibility for artificial intelligence to become a legal subject.

Conclusion
Artificial intelligence cannot challenge human's status as the original legal subject, which is actually the position of legal anthropocentrism.Anthropocentrism is the principle and bottom line that must be adhered to when thinking about the subject status of artificial intelligence law.It is anthropocentrism that enables human beings to get rid of the dilemma that life is meaningless and existence has no basis.Anthropocentrism in the field of metaphysics maximizes the consolidation of human beings as doctrinal matrix, standing in the world and taking themselves as the source of meaning, making reason and free will become the authority.This significance is not only for the dignity and value of individuals.At the same time, it also provides meaning for the social order and political process, so that we can believe in the value and significance of the legal system as a human social governance scheme.In this process, we take ourselves as the subject to achieve the beneficial coordination between personality and commonality, material and spirit, and become the creator of the law, but also obey the law created by ourselves, realize the organic unity of self with self, with others and with the whole.More importantly, only by emphasizing anthropocentrism can we stay alert to the dissolution of human subjectivity in the era of artificial intelligence in the development of science and technology.However, this does not mean that we shall emphasize the supremacy of anthropocentrism in the subject status of artificial intelligence law.When looking into the development prospect of modern technology and its interaction with law, the legal subject must get rid of the bondage of narrow anthropocentrism, because if the narrow anthropocentrism is used as an excuse to arbitrarily deny artificial intelligence from the value of the legal subject, it may be straightfor-ward, but it will impose the shackle on the development of law and even human development.The historical experience of legal person as a derivative legal subject tells us that although legal fiction technology is a utilitarian approach adopted by positive law to solve practical legal problems, it still hides the consideration of legal value, that is, all rational considerations are focused on the ultimate goal of whether it is conducive to the fundamental interests of people.In fact, this is also a position of legal anthropocentrism.First, it is reflected in that the motivation of legal fiction is to better improve the efficiency of law and serve the development needs of human society.Laws expand or narrow the scope of the subject based on specific purposes, indicating that the reason for obtaining the qualification of the subject of law lies in whether there are legally recognized interests to be protected.From this point of view, legal fiction originates from human's own needs, and the development of the legal subject must be beneficial to human itself.Only on this premise can there be the necessity of legal fiction.As Schmidt, a German scholar, stressed: Law exists for people, not for factories or football associations, and not for the state.But this does not mean that we can handle everything without a legal person.On the contrary, we need an organization that transcends individuals and has rights and obligations for the purpose of serving people (Schmidt, 2003).It can be seen from the fiction of legal subject of legal person that the technology of legal fiction is only carried out at the level of specific legal practice, and does not challenge the core value of natural person as the legal subject.The technology itself is not to construct an independent new value system, but to realize the greater value of human as the original legal subject through the construction of derivative subject.From the perspective of legal technology itself, the standard for artificial intelligence to become a legal subject does not lie in the pure legal technology itself, but in the value logic behind the legal technology.The key point of whether artificial intelligence can become a legal subject is not whether artificial intelligence can be granted rights and can undertake obligations, but whether artificial intelligence should be granted rights and undertake obligations.When discussing the latter, the key question is whether it is in line with the ultimate goal of human interest in the long run, while the core value of human as the subject of law should not be challenged.

(
self-consciousness) as the starting point of cognition, and placed people in the position of cognitive subject, highlighting the subjectivity of people in cognitive relations.Descartes elevated the cognitive subject and the thinking subject to a new level, and made it clear that the phenomenal world is the result of the inference of the subject's self-cognition.It confirms the existence of "I" as a matrix subject, and then turns the external world into an object represented, recognized lationship between the persons as the legal subject, and the legal relationship between the subject of law and the object of law.With the development of positivism jurisprudence, the meaning of the legal subject in such construction is initially abstracted on the basis of concrete people in reality.The universality of law requires further abstraction of the subject of law into a broader person, a person who removes the individual characteristics of all concrete people.Based The concept of right capacity is a deep abstraction based on legal personality, and the important symbol of it is depersonalization.As part of the subject category of law, right capacity is the legal status obtained through the provisions of substantive law.The concept of capacity for rights was first proposed in the Fragments of the Civil Law System of the City (1789).In the early 19th century, Thibaut was the first to elaborate on the capacity of rights in theory.He believed that the capacity of rights was divided into the capacity of natural rights and civil rights.The former contained all physical attributes that could produce a specific legal relationship, that is, the natural identity, which was the basis for biological man to become a legal person.The latter was what was known in Roman law as Caput or Status, the external constraint that positive law imposed on the legal person.The German Civil Code in 1896 and the Swiss Civil Code in 1907 specifically created the capacity of rights in the substantive law.Compared with the concept of legal personality, the capacity of rights further abstracted the legal "person", and the corresponding qualification of legal subject was fixed from the abstract representation of "person" to the concept of the rights that "person" has or is able to have.The capacity of rights means "enough to have rights or assume obligations".Only the entity with the capacity of rights can become the subject of law.The concept of right capacity replaces the concept of personality.By replacing the concept of "qualification" which needs to be recognized or endowed by others with the concept of "ability" in the context of objective facts, the natural person's birth with the right capacity becomes a legal event established by the legal subject.The sovereign does not need to recognize it specifically, which makes the positive law more thoroughly reflect the thought of natural rights.
to replace human work.The main way of simulation is through computer programs (or "algorithms").Third, "independent thinking".At present, artificial intelligence has the ability of deep learning, self-learning and independent judgment and action.Fourth, "multiple forms".Artificial intelligence can be either a humanoid robot or a computer device containing intelligent systems, intelligence is the product of human creation, which indicates that artificial intelligence is not natural generation, but some extension of human.Second, "simulated intelligence".Artificial intelligence can simulate human intelligence in Z. F. Wen, D. Y. Tong DOI: 10.4236/blr.2023.14100480 Beijing Law Review some aspects