release № / 01 апр 2026 Скачать PDF статьи

DIGITAL AFTERLIFE AND THE ILLUSION OF CONTINUITY: LEGAL AND ETHICAL ASPECTS OF POST-MORTEM DIGITAL SIMULATION

UDC: 34:004.8:340.12

DIGITAL AFTERLIFE AND THE ILLUSION OF CONTINUITY: LEGAL AND ETHICAL ASPECTS OF POST-MORTEM DIGITAL SIMULATION

Kildeev Adel

Zhdanov A. Leningrad State University (1990)

Philosophiæ Doctor in Law

ORCID: 0009-0001-2211-4944

Abstract: The article is devoted to the legal and philosophical analysis of the phenomenon of the digital afterlife — the creation and use of algorithmic systems that imitate the personhood of a deceased person on the basis of digital data left by him during his lifetime.

The author proceeds from the thesis of the fundamental difference between the personhood as a bearer of consciousness and the algorithmic model as an instrumental construct devoid of intentionality and subjectivity. It is shown that the widespread rhetoric of “digital immortality” is based on the substitution of ontological categories and forms the illusion of the continuation of existence where there is only a statistical reconstruction of speech and behavioral patterns.

Based on the concepts of I. Kant, G.W.F. Hegel and N.A. Berdyaev, the impossibility of the subject in the absence of a self-identical bearer of freedom and action is revealed. The concept of simulacrum is introduced as a sign form of subjectivity without a subject. The analysis of the artistic models of N.V. Gogol and V.O. Pelevin demonstrates the cultural and anthropological dimension of this problem.

It is concluded that the digital simulation of the individual affects the principle of human dignity and requires special legal regulation, while in digital communication a person acts, and artificial intelligence remains a tool for the implementation of his will.

Keywords: digital afterlife, post-death personhood rights, AI simulation, artificial intelligence; legal personhood; large language models; philosophy of law; simulacrum; subject; the transcendental unity of apperception; recognition; human dignity; digital personhood simulation; responsibility; instrumentum vocale.

Use of AI tools in the preparation of the article: In the preparation of this article, artificial intelligence tools — the ChatGPT (OpenAI, 2025) and Gemini (Google DeepMind, 2025) models — were used exclusively for technical and editorial support: language optimization, link structuring, and terminology unification. All scientific conclusions, interpretations and analytical generalizations belong to the author and reflect the results of his own research work.

Chapter 1. Introduction.

For the first time in history, the digital environment has created the technical ability to simulate the presence of a deceased person. On the basis of texts, images, audio and video recordings, algorithmic models are formed that are capable of reproducing speech and behavioral patterns of the individual. In public discourse, this phenomenon has been called “digital immortality”, which already at the level of terminology contains a significant categorical substitution.

It is not a question of continuing the existence of the individual, but of creating a statistical reconstruction of it, devoid of consciousness, will, and intentionality. An algorithmic system does not have subjectivity and cannot be a bearer of rights and obligations. Nevertheless, in social and economic circulation, such a system begins to function as a quasi-personhood, which gives rise to fundamentally new legal and ethical collisions.

The problem of the digital afterlife lies at the intersection of several areas:

  • the conflict between the right of the individual and the protection of human dignity and the principle of civil circulation,
  • between the right to data and digital inheritance and inheritance law,
  • intellectual property,
  • regulation of artificial intelligence.

Unlike traditional forms of memory of the deceased — text, photo, or audio recording — algorithmic simulation creates the effect of dialogue and thereby creates an experience of continued presence in third parties. A new legal reality arises: an object that is not a person, but is capable of reproducing its social role, a simulacrum of personhood. Its definition and regulation requires normative legal qualification.

This requires answering fundamental questions:

  • whether it is permissible to create a digital simulation of a person without his lifetime consent;
  • who owns the rights to such a simulation;
  • where is the line between the memory of the deceased and the exploitation of his identity;
  • whether human dignity can have a posthumous legal dimension.

Thus, the digital afterlife is not a technological, but an ontological and legal problem. It affects the basic categories of personhood, identity, and responsibility, and requires the formation of a new normative model that can protect a person from becoming an algorithmic object after death.

This study continues the series of the author’s works devoted to the legal and ontological qualification of artificial intelligence.[1] In these articles, the author examined in detail the legal status of AI, the ethical standards of AI, aspects of AI and copyright, numerous court cases against AI companies and their litigation with their employees and between the AI companies themselves, and finally, the legal theoretical and philosophical basis of AI, its international legal regulation and proper place in world technologies and economy, the fundamental differences between LLM and artificial intelligence. The philosophical, legal and historical prerequisites of the concept of artificial intelligence, including the concept of instrumentum vocale and the model of the “slave of the lamp” in the Roman legal tradition, are considered.

The development of generative models for the first time created a technological opportunity for a stable digital simulation of an individual, which does not have a formed legal qualification. There are separate studies in three unrelated blocks: 1. digital heritage and the digital afterlife; 2. Personal data rights and digital identity; 3. AI Ethics and LLM. But there is no comprehensive analysis of the generative simulation of the individual as an independent legal object.

Object and subject of research

The object is social relations arising in connection with the technology of generative simulation of personhood.

The subject is the legal and ontological nature of digital emulation of personhood, created on the basis of large language models.

Purpose of the study

The purpose of the study is to form a holistic legal qualification of the technology of generative simulation of the individual and to determine its place in the system of personal non-property rights.

Objectives of the study

  1. To identify the difference between generative personhood simulation and:

    • Digital Archive
    • avatar
    • user profile.
  2. To determine the technological prerequisites for the emergence of this phenomenon.
  3. To analyze the applicability of existing legal regimes.
  4. To investigate the risks of post-mortem digital representation.
  5. Formulate proposals for legal qualification.

Research methodology

The methodological basis of the research consists of formal legal, comparative-legal, interdisciplinary, and ontological analysis.

The scientific novelty of the research lies in the fact that for the first time generative simulation of personhood is considered not as a kind of data processing or digital memory, but as an independent legal phenomenon, an independent object of legal regulation.

Chapter 2. Historical and Technological Genesis of Digital Personhood Simulation

2.1. Pre-digital forms of the presence of the deceased as a cultural and legal phenomenon

In all known cultures, the memory of the deceased was provided through symbolic forms of representation, such as text, image, ritual, or oral tradition. These forms could be socially significant and have a high degree of sacralization, but they did not imply the reproduction of individual will, speech activity and behavioral variability of the individual. The image of the deceased was preserved as an object of memory, but not as a subject of communication.

In the Roman tradition, the function of representing ancestors was performed by the imagines maiorum — wax masks worn during funeral processions and used in public ceremonies.[2] They provided a symbolic presence of the dead in political and family life, but they did not reproduce their individual consciousness and were not endowed with the ability to express themselves autonomously.

A similar logic can be traced in the Confucian tradition of the cult of ancestors, where ritual communication with the dead was carried out through strictly regulated forms of veneration.[3] The ancestor was the object of a sacred relationship, but not the subject of dialogue.

In the biblical tradition, this boundary has a normative character. The Old Testament contains a direct ban on attempts to establish arbitrary contact with the world of the dead (Deut. 18:10-12)[4], which testifies to the fundamental inadmissibility of non-ritualized reproduction of their “voice”.

In the Vedic tradition, the cult of ancestors (śrāddha) is strictly ritualized and aims to maintain cosmic and social order (Ṛgveda X.15; Manusmṛti III.122–125).[5] [6]

In Islamic legal culture, the sacralization of the human person after death is expressed, in particular, in the special protection of the bodily integrity of the deceased and the prohibition of arbitrary construction of his image.[7]

In all these cases, violation of the established forms of treatment of the memory of the dead was considered a grave religious and legal offense. Thus, an archaic model of posthumous protection of the individual was formed.[8] [9]

Consequently, before the advent of digital technologies, all cultural systems retained a fundamental boundary between the memory of the individual and its acting communicative substitute.

2.2. Early digital forms of the presence of the deceased: digital archive, memorialization of accounts and avatars

The development of the digital environment led to the emergence of new forms of preserving the posthumous presence of the individual, but at the initial stage they did not go beyond the traditional model of representation and did not affect the communicative autonomy of the subject. It was about recording, storing and institutionalizing a digital footprint created during the user’s lifetime.

The simplest form is a digital archive, which is a structured set of texts, images, audio and video materials. Such data arrays do not have the ability to generate new content and therefore retain the status of an object, not a subject of communication. Their legal regime is determined by the norms of inheritance law, legislation on personal data and provisions on the results of intellectual activity.[10] [11]

The legal qualification of the digital data of the deceased has been developed in judicial practice. The decision of the Federal Court of Germany of July 12, 2018 in case III ZR 183/17, in which the account of the user of the social network was recognized as part of the estate and subject to transfer to heirs by way of universal succession, is precedential.[12] The court pointed out that the digital profile is a set of contractual rights and information content and cannot be considered as an extension of the identity of the deceased.

A similar approach is enshrined in American law under the Revised Uniform Fiduciary Access to Digital Assets Act (2015), which granted inheritance managers the right to access the digital assets of the deceased subject to established procedures.[13] Thus, the digital data were finally included in the sphere of inheritance succession.

The development of user services has led to the formation of a mode of account memorialization. In the social network Facebook, the status of memorialized account has been introduced, in which the profile is saved on the network as memory space, but loses the ability to act independently, and its management is transferred to a trusted person appointed during his lifetime (legacy contact).[14] Similar mechanisms are provided in Google (Inactive Account Manager) and Instagram services.[15]

The fundamental circumstance is that in these models, the account:

– does not generate new messages;

– does not enter into a dialogue;

– does not reproduce the behavioral variability of the personhood.

Its function is limited to preserving previously created content and providing collective digital memory.

A special place among the early forms of digital representation of the individual is occupied by avatars used in virtual worlds and virtual reality systems. In environments such as Second Life or VRChat, an avatar is a graphical representation of the user that can imitate gestures, movement in space, and other elements of bodily behavior.[16] [17]

Despite the high level of visual individualization, the avatar does not have communicative autonomy. His actions are completely determined by the will of a live user or a predetermined algorithm. After the death of the user, the avatar loses its active function and can only exist as an element of the digital archive. Its legal regime is determined by the user agreement of the platform and the rules on intellectual property.[18]

Thus, even the most complex forms of digital visualization of a person do not create the effect of his independent presence in social communication.

In the scientific literature, these processes are considered within the framework of the concept of digital heritage and posthumous privacy, where the main problem is the distribution of access to data and their legal fate after the death of the subject.[19] [20]

Consequently, early digital forms of postmortem presence are characterized by the preservation of an information trace without reproducing the communicative function of the individual. Their legal regime fits into the existing structures of inheritance, contract and information law.

A qualitative break with this model occurs only with the advent of technologies capable of generating new statements that imitate the individual style and behavioral reactions of the deceased. It is this transition that transforms digital representation from the form of an archive into a form of simulation of the individual and takes it beyond the boundaries of traditional legal regimes.

Chapter 3. Generative Simulation of Personhood as a Technological and Legal Reality

3.1. Meta’s Patent as a Fixation of the Technological Possibility of Generative Personhood Simulation

The transition from the digital representation of a person to its generative simulation ceases to be a theoretical construct at the moment when the corresponding technology receives a patent form. The Meta Social networking system that simulates a user using a language model patent application captures a reproducible technical solution designed to function in the social network infrastructure.[21]

The specification of the invention proposes an architecture in which a pre-trained large language model is adapted based on the historical data of a particular user. Publications, comments, reactions, as well as behavioral scenarios for interacting with content are used as a training array. These data are transformed into an individualized probabilistic model of speech and communicative behavior.

The system generates a custom training dataset, analyzes the incoming content, generates a response, and places it in the digital environment on behalf of the user. In this way, it is not the text that is reproduced as a static object, but the structure of social action, i.e., communication addressed to another participant in the network.

It is of fundamental importance that the description of the invention expressly provides for the operation of the system in the absence of the user, including cases of his death.[22] The technology is focused on the continuation of communicative activity when the physical existence of the subject ceases.

Legally significant features of the technical solution under consideration are:

– personalization of the model formed on the basis of the data of a particular person;

– autonomy of communication carried out without the current expression of the will of the subject;

– inclusion of the results of the system’s functioning in real social interaction;

– independence of functioning from the biological time of human life.

Taken together, these features mean the reproduction of the communicative function of the individual.

The patent form of fixation gives this technology an independent legal significance. The described solution is considered as an object intended for implementation, licensing and economic use in a platform environment. Thus, the reproduction of an individual model of behavior is included in civil circulation.

The technical possibility of separating the communicative function of a person from its biological bearer of consciousness and its autonomous functioning in the social environment is documented, which in business and technological publications is directly interpreted as the creation of a digital substitute for the user. [23] [24]

3.2. Functional Features of Generative Simulation of Personhood as an Object of Legal Qualification

Legal qualification of generative simulation of personhood is possible when identifying its functional characteristics that are important for law.

The first feature is the generation of previously non-existent content. The algorithmic system forms new statements that respond to the changing communicative context.

The second feature is personalization. It is not an abstract model of behavior that is reproduced, but an individual identity that allows third parties to perceive the result of the functioning of the system as the actions of a particular person.

The third feature is the autonomy of functioning. Speech behavior is separated from the physical bearer of consciousness and acquires the ability to exist independently.

The fourth feature is social integration. Generative simulation is included in real communication and influences the behavior of other participants.

The fifth feature is the temporary unlimited functioning.

The totality of these characteristics means the reproduction of the communicative function of a person regardless of his physical existence.

This phenomenon cannot be adequately described within the framework of existing legal categories. It is not reduced to personal data, since the object of regulation is not information, but a model of behavior; is not the result of intellectual activity, since the method of participation in communication is reproduced; does not fit into the inheritance legal structure, since we are not talking about the transfer of an array of data, but about the continuation of social activity. Thus, the generative simulation of personhood has the features of an independent object of legal qualification.

3.3. Differentiation of Generative Simulation of Personhood from Related Legal Constructions

Generative simulation of personhood does not coincide with any of the existing legal categories.

  1. It is not personal data, since it reproduces communicative activity, and not information about a person.
  2. It does not belong to the results of intellectual activity, since it models identity, and does not create a work.
  3. It does not form an object of inheritance succession, since there is no transfer of the existing property complex.
  4. It does not coincide with the memorial representation, which is static.
  5. It is not an avatar, the functioning of which depends entirely on the will of the user.

The differentiating criterion in all cases is the presence of a reproducible communicative function of the individual.

If in traditional legal constructions the object of regulation is data, image, work or property complex, then in the case under consideration a social action is reproduced — participation in communication on behalf of a particular person.

This confirms the need for an independent legal qualification of generative simulation of personhood.

3.4. Conclusions

The patent fixation of the technology confirms the possibility of autonomous reproduction of the communicative function of the individual.

Generative Simulation:

– generates new content;

– reproduces individual identity;

– functions in social communication;

– is not limited by the lifetime of the subject.

These features allow us to consider it as a new object of legal regulation. The object of protection is the reproducible identity of the person in communication.

Chapter 4. Artificial Intelligence as an Imaginary Subject: Legal Requalification of LLM and the Status of a Simulacrum

The initial doctrinal formulation of the question of the legal personhood of artificial intelligence is based on the criterion of reasonableness as the basis of legal personhood. In the most concentrated form, this position is expressed by Lawrence Solum:

“If the entity can reason, it can be a legal person.”[25]

The legal meaning of this statement is that legal personhood is derived from the ability to rational decision-making. However, the applicability of this criterion presupposes a preliminary establishment that the object in question has an autonomous ability to set goals and acts as a participant in legal circulation. Large language models have neither a sphere of goals of their own, nor the ability to initiate legally significant actions. Their functioning is completely determined by the user’s request and the operator’s infrastructure. Therefore, Solum’s thesis refers to a hypothetical autonomous agent and does not apply to the LLM as an instrumental text generation system. Already at this level, the discussion about legal personhood loses its subject, since the object is qualified erroneously.

A similar substitution occurs in the corporate model proposed by Sean Bayern, who states:

“there is no conceptual barrier to recognizing an autonomous system as a legal entity.”[26]

This thesis assumes that the legal personhood of a legal entity is a purely normative construct and can be extended to any sufficiently complex system. However, the legal personhood of a corporation is based on two legal elements: property isolation and organizational structure, through which the legally relevant will is formed. The LLM has neither its own property nor organs. It does not make decisions in the legal sense but generates text sequences within a given architecture. Consequently, the recognition of its legal personhood does not create a new center of responsibility. Moreover, text generation does not constitute a legally relevant act. A legally significant action is performed by the subject using the result of generation. Thus, the LLM does not act as an agent, but as a means of creating a sign form that imitates communicative behavior.

In the criminal law doctrine, Gabriel Hallevi formulates the thesis of the functional equivalent of guilt:

“AI can fulfill the requirements of mens rea in a functional sense.”[27]

The legal significance of this thesis lies in the reduction of guilt to the cognitive element. However, guilt in criminal law includes not only an intellectual, but also a volitional element, as well as the possibility of personal imputation and application of punishment to the subject. The algorithm is not the addressee of criminal repression and cannot be sanctioned. Consequently, the transfer of the category of mens rea to it destroys the very structure of criminal responsibility. In the event of harm, the person using the system remains the subject, since it is he who decides to use it and bears the risk of such use.

The modern theory of autonomous social systems in the person of Günther Teubner allows for the possibility of the emergence of “electronic agents” functioning in law as participants in communication.[28] However, in this model, the subject is not a technical system as such, but a communicative structure that has the ability to self-refer and reproduce its own operations. The LLM has neither self-reference nor its own normative program; it reproduces probabilistic language sequences within the limits of external control. Consequently, even within the framework of the system theory of law, it cannot be recognized as an autonomous participant in legal communication.

Ugo Pagallo, analyzing the possibility of a new form of legal personhood for robots, connects it with their participation in social interactions.[29] However, participation in communication as such is not a legal criterion of the subject. The law does not qualify the form of behavior, but the presence of a bearer of rights and obligations.

Thus, all the main doctrinal arguments in favor of the legal personhood of artificial intelligence are based on the identification of an autonomous agent with a text generation system. In reality, an LLM is not artificial intelligence in the sense of an actor. It is a data processing tool that creates a simulation of speech behavior.

The legal nature of a simulacrum is determined by its object character. It can be:

– the object of intellectual rights,

– the object of regulation in the field of data protection,

– a source of increased risk,

but not the subject of a legal relationship.

Law knows only two basic categories – subject and object. An object that reproduces the form of action does not acquire legal personhood, since legal personhood is determined by the ability to be a bearer of rights and obligations and to bear legal liability.[30]

In this regard, the qualification of an LLM as an instrumentum vocale fixes its actual position in the legal system: it is a technical means by which the legal subject performs legally significant actions. All legal consequences of its functioning relate to the persons who determine the purposes of its application and control its life cycle.

It follows that the so-called responsibility gap does not arise from the autonomy of the system, but from the erroneous ontological qualification of it as an acting agent. In reality, the action is performed by the subject using the tool, while the simulation of communication remains the object of legal regulation.

Positive law confirms this qualification. Regulation (EU) 2024/1689 imposes obligations solely on providers and users of artificial intelligence systems and does not provide for their legal personhood. Thus, the legislator fixes the instrumental status of the relevant technologies.

Consequently, the debate about the legal personhood of artificial intelligence in relation to LLMs is a categorical fallacy. The object of legal analysis is not an autonomous subject, but a technical tool that creates a simulation of communicative behavior.

It is significant that the positive law of the European Union, taking into account the arguments about the responsibility gap, has not followed the path of creating a new legal personhood for artificial intelligence. Regulation (EU) 2024/1689 regulates in such a way that the only subjects of duties and responsibilities are persons involved in the life cycle of an artificial intelligence system. These include suppliers, authorized representatives, importers, distributors and users artificial intelligence systems (Article 3).[31] They are responsible for system compliance, risk management, technical documentation, transparency, and human oversight (Articles 8 to 29).

At the same time, an AI system is defined as a machine system designed to operate with varying degrees of autonomy (Article 3 (1)), and in the entire structure of the Regulation appears exclusively as an object in respect of which safety requirements and restrictions on putting into circulation and use are established. It is not vested with rights, does not bear obligations and cannot act as the addressee of sanctions. The addressees of administrative measures and fines are exclusively economic entities (Article 99).

Of particular importance is the requirement of human oversight, which is enshrined among the mandatory requirements for high-risk systems (Article 14). This norm directly proceeds from the fact that a legally significant decision must remain within the control of a person and can be changed or canceled by him. Thus, the possibility of recognizing the system as an autonomous bearer of consciousness of legally relevant will is excluded.

The normative structure of the act demonstrates that the problem of responsibility gap is resolved by establishing a continuous chain of duties and responsibilities of entities that put the system into circulation and use it in professional activities. Consequently, the gap in responsibility arises not from the autonomy of technology as such, but from its erroneous qualification as an acting agent. In the legal model of AI Act, the artificial intelligence system acts as a source of risk and an object of regulation, while the legal subject remains the bearer of duties and responsibilities.

Normative conclusion

Regulation (EU) 2024/1689 structurally excludes the possibility of recognizing artificial intelligence as a legal entity, since:

– only participants in economic turnover are subjects of obligations;

– sanctions are addressed exclusively to them;

– A legally significant decision must remain under human control.

Consequently, the responsibility gap is eliminated by means of the existing law and does not require the introduction of a new legal personhood.

Final Conclusions and Legal Definition of Simulacrum

The analysis shows that the doctrinal constructions substantiating the possibility of recognizing artificial intelligence as a legal subject are based on the premise of the presence of an autonomous bearer of consciousness of legally relevant behavior. Large language models, however, have neither a sphere of goals of their own, nor the ability to make decisions of legal significance. Their functioning is fully determined by external goal-setting and is implemented within the framework of the infrastructure belonging to the subjects of law. Consequently, they cannot be considered as participants in legal relations.

The analogy with a legal entity turns out to be untenable due to the lack of property isolation and organ will; the concept of functional compliance with the subjective side of the crime due to the impossibility of personal imputation and the addressability of the sanction; The argument of responsibility gap is due to the fact that a legally significant action in all cases remains in the sphere of a person’s decision to create, put into circulation and use a system. The positive law of the European Union, which enshrined a regulatory model in which all obligations and sanctions are addressed exclusively to participants in economic turnover, normatively confirms the instrumental nature of artificial intelligence.

Thus, legal analysis leads to the need to distinguish three levels, which are often identified in modern discussion:

  1. hypothetical autonomous artificial intelligence as a possible subject;
  2. large language models as a data processing tool;
  3. the communicative form they create.

The object of this study is the third level, the sign structure that reproduces the form of the subject’s speech behavior.

In the legal sense, this structure is subject to qualification as a simulacrum.

In this study, a simulacrum is defined as a sign form created by a technical system that imitates the communicative action of the subject, but does not have an autonomous will, property isolation and the ability to bear legal responsibility.

Such a definition has direct legal consequences. A simulacrum can be:

– the object of intellectual rights;

– the object of regulation in the field of personal data protection;

– a source of increased risk;

– a means of performing legally significant actions by a legal subject.

However, it cannot be:

– the bearer of subjective rights;

– a participant in a legal relationship;

– the subject of tort or criminal liability.

Consequently, the reproduction of the form of communicative behavior does not create a new participant in the legal turnover. It is not the external structure of the action that has legal significance, but the presence of a bearer of legally relevant will and responsibility.

In this regard, large language models are subject to qualification as instrumentum vocale – a technical means by which the legal subject performs legally significant actions, while the simulacra created by them are objects of legal regulation.

An object that functionally reproduces the features of the subject’s speech and cognitive behavior does not acquire legal personhood, since it is not the form of action that has legal significance, but the presence of an autonomous will, property isolation and the ability to bear responsibility. All these elements in relation to text generation systems belong to the persons who determine the goals of their operation.

Thus, the problem of the legal personhood of artificial intelligence as applied to large language models (LLMs) is not an unresolved issue of law, but a consequence of a categorical error in the definition of the object.

Chapter 5. Simulacrum and the Problem of the Subject: Philosophical Foundations of Legal Qualification

5.1. Subject as a condition for attribution of an action

The legal qualification of large language models as a tool and the sign structures they create as an object of law requires philosophical substantiation. Law does not construct the subject arbitrarily but proceeds from the possibility of attributing an action to a certain bearer of will. Consequently, the question of the subject precedes the legal qualification.

Modern functionalist theories of consciousness associate subjectivity with a certain level of information processing complexity. Classical philosophy, on the contrary, considers the subject as a condition for the possibility of experience, freedom and responsibility.

5.2. I. Kant: Transcendental Unity of Apperception

Kant’s formula

“The thought ‘I think’ must be able to accompany all my ideas.”[32]

means that the subject is not the result of synthesis but acts as its condition. The unity of self-consciousness makes possible the continuity of experience and the attribution of action.

An algorithmic system is capable of syntactic and semantic synthesis, but each of its states arises anew and does not belong to the same “I”. It has no internal history and cannot be the bearer of the act. Consequently, the form of thinking is reproduced without the subject of thought.

5.3. G.W.F. Hegel: Recognition, Capacity for Negation and Freedom

Hegel defines the subject as self-consciousness existing in the space of recognition:

“Self-consciousness is in itself and for itself insofar as it exists for another self-consciousness.”[33]

The subject is the one who is able to go beyond the limits of the given, to deny the immediate, and to act from freedom.

An algorithmic system does not have negativity: it cannot but be what it is. It is capable of reproducing the form of dialogue but does not participate in the dialectic of recognition.

5.4. The Limit of Functionalism

Functionalism proceeds from the quantitative difference between man and machine. However, in Kant and Hegel, the subject is defined structurally, through self-identity and freedom. No computational procedure creates an action attribution point.

5.5. D. Dennett: Intentional Stance and the Mechanism of Simulacrum Emergence

Dennett proposes to consider the subject as the result of interpretation:

“A system is an intentional system insofar as its behavior can be predicted by attributing beliefs and desires to it.”[34]

The language model fits this scheme perfectly, since its behavior is coherent and rationally interpretable.

However, interpretation does not create a transcendental unity of self-consciousness and does not generate freedom. Consequently, the intentional position does not describe the subject, but the mechanism of the emergence of the simulacrum.

5.6. A.F. Losev: Mythological Form of Personhood

Losev shows that a sign can be experienced as a reality of personhood:

“Myth is not fiction, but reality itself, given as personal being.”[35]

The digital “personhood” has just such a structure of presence, but in the absence of a self-identical carrier, the sign remains an empty form.

5.7. N.A. Berdyaev: Freedom as a Criterion of Personhood

Berdyaev defines personhood through freedom:

“Personhood is freedom and a creative act.”[36]

An algorithm belongs entirely to the sphere of necessity and therefore cannot be a source of action.

5.8. N.V. Gogol: Three Models of the Simulacrum

“Dead Souls”: A Legal Entity Without a Carrier

The name of the deceased continues to function as a legal reality.[37] A sign acts as a subject in the absence of a subject.

“The Overcoat”: Reduction of Personhood to the Function of Text

Akaky Akakievich exists as a pure function of copying.[38] Personhood coincides with the operation of reproducing a sign.

The Nose: Recognition Without Self-Awareness

Major Kovalev’s nose exists as an independent social figure with a higher status than its previous owner.[39] He is recognized, occupies a place in the hierarchy and enters into communication as an autonomous person. Here subjectivity is completely transferred to the sphere of recognition, while there is no self-identical carrier.

5.9. V. Pelevin: Interface Model of the Subject and the Ontology of the Simulacrum

In the late novels of V. Pelevin, the subject finally turns into a discursive function.

In iPhuck 10, the narration is conducted on behalf of an algorithm that has memory, style and self-reference, but exists exclusively as a software construct.[40]

In Transhumanism Inc., personhood breaks down into digital shells and becomes a configuration of data.[41]

In KGBT+, the subject coincides with its medial image and exists as a function of affect production.[42]

In all these cases, discourse has a structure of subjectivity in the absence of its carrier. This is the phenomenology of the simulacrum.

5.10. Philosophical Summary

The subject is impossible without:

– transcendental unity of self-consciousness;

– freedom;

– recognition;

– creative source of action.

Language models reproduce the form of these manifestations without their ontological foundation. A simulacrum is a form of subjectivity without a subject.

Philosophical analysis thereby confirms the legal qualification of artificial intelligence as an instrumentum vocale – a means of action of the legal subject.

Chapter 6. Conclusion: the limit of simulation and the dignity of the person

The problem of the legal personhood of artificial intelligence has in fact turned out to be not a question of the future of technology, but a question of the limits of the human. Throughout the study, the initial confusion of levels was consistently eliminated, in which the technical ability to generate discourse was taken for the presence of a subject, functional coherence for the transcendental unity of apperception, and social interpretation for freedom. As a result, what belongs to the sphere of means was endowed with the status of the bearer of action.

Legal analysis has shown that in all cases where an artificial intelligence system operates, a person acts – the one who determines the goals of its functioning, puts it into circulation and extracts the result of its application. Philosophical analysis has revealed a deeper basis for this conclusion: the subject is impossible without the self-identical unity of experience, without freedom as the ability to initiate action, and without recognition, which presupposes the meeting of two wills. No computational procedure creates these conditions, no matter how complex it may be.

Thus, a special phenomenon arises in the space of digital communication — a form of subjectivity without a subject. She has a voice, style, memory, the ability to dialogue, and social recognition. It can be interpreted as a person and even recognized in this function. However, its existence is limited to the sign structure, in which there is no internal source of the act. This is a simulacrum – not an illusion or a perceptual error, but a new form of an object functioning as a subject.

It is here that law encounters an anthropological limit. As long as we are talking about a technical tool, the problem remains in the field of responsibility, data circulation and risk management. But the moment the simulacrum begins to reproduce a concrete human personhood—its voice, its image, its way of speaking and thinking—a question arises that cannot be solved by technology alone.

Personhood is not reduced to a set of texts, images and behavioral patterns. It is the source of action, inner self-identity and freedom. Therefore, its digital reproduction is not its continuation. It is always a construct created by others and used by others.

In this sense, the idea of “dead souls” takes on a new meaning in the digital age. The existence of personhood as a pure sign function, which retains legal and social effectiveness in the absence of a living carrier, comes into direct contradiction with the personalistic tradition of European culture, including Christian anthropology, in which man is thought of as a unique and indivisible unity with dignity that is not reducible to its manifestations. A name cannot be separated from a person and turned into an object of circulation without losing the very concept of a person.

That is why the protection of the memory of the deceased has always occupied a special place in law. It testifies to the fact that the dignity of the person does not disappear along with his biological existence. A digital simulation that creates the illusion of continued presence breaks down this boundary, turning a person into a functional interface.

At this point, the legal, moral and cultural dimensions coincide. The ban on profanation of the memory of the dead exists not as an archaic ritual, but as a form of protection of human subjectivity from becoming a thing. Technologies that make it possible to reproduce personhood as a discourse call into question this limit and thus require a new understanding of it.

Russian classical literature has already described both possibilities of human existence. In the images of Pushkin, Dostoevsky and Tolstoy, personhood is revealed as inner freedom and a source of action, irreducible to function. In Gogol, on the contrary, a world arises in which a name, rank, text and even a look begin to live their own life, separating from their bearer. Modern digital reality only gives this Gogol’s intuition a technical form.

Therefore, the question of simulacra is not a question of technology. It is a question of whether man will survive as a subject in a world where everything can be reproduced as a sign.

A normative limit follows from the analysis: digital reproduction of a person is permissible only where it is based on the free and clearly expressed will of the person himself. In all other cases, it must be regarded as an unacceptable form of treatment of human dignity.

This conclusion is not an external moral addition to the legal construction. It flows out of it. Legal personhood is possible only where there is a bearer of freedom and responsibility. Where there is none, there is only a means.

In digital communication it is always the human who acts-the simulacrum merely speaks. The preservation of this distinction becomes a condition for the preservation of the very concept of personhood. The boundary between the subject and the simulacrum is the boundary between man and his transformation into a thing; Law, philosophy and culture coincide in the demand to preserve this boundary.

And the Artificial Intelligence should be the slave of the lamp, and the Man should be his master, and not vice versa.

References

        1. Kildeev A.Kh. Artificial Intelligence as a Servant of Man or Vice Versa: Legal Aspects, Scientific Platform “21st Century” Journal, # 1, April 2025, www.episs.ru;
        2. Kildeev A. Kh. Ethical Application of Artificial Intelligence: A Global Review of Principles, Practices, Regulations and Implications, Journal SCIENTIFIC PLATFORM “21st CENTURY”, # 2, May 2025, www.episs.ru;
        3. Kildeev A.Kh. Artificial Intelligence and Copyright: Legal Aspects, Scientific Platform “21st Century” Journal, # 3, June 2025, www.episs.ru;
        4. Kildeev A.Kh. Disney and Warner Brothers Discovery Lawsuits v. Midjourney and Matthew and Maria Raine v. OpenAI and Sam Altman and Their Implications for the Development of the Artificial Intelligence Industry, 21st Century Science Platform, #4, July 2025, www.episs.ru;
        5. Kildeev A.Kh. X.AI Corp-X.AI LLC lawsuits against Xuechen Li and X.AI Corp.-X.AI LLC against OpenAI and implications for the artificial intelligence industry, 21st Century Scientific Platform Journal, #5, August 2025, www.episs.ru;
        6. Kildeev A.Kh. Artificial Intelligence-Slave Lamp, Scientific Platform “21st Century” Journal, # 6, September 2025, www.episs.ru;
        7. Kildeev A.KH. LLM is not Artificial Intelligence. Scientific Platform “21 CENTURY” journal, # 7, October 2025, www.episs.ru.
        8. Polybius. Universal History / transl. from ancient Greek. – Moscow: Nauka, 1986.

          (Description of the Imagines Maiorum, Book II. VI).

        9. Li ji (The Book of Rites) / transl. from Chinese – Moscow: Oriental Literature, 1999.
        10. The Bible. Books of the Holy Scriptures of the Old and New Testaments. Moscow: Russian Bible Society, 2011.
        11. Ṛgveda / transl. from Sanskrit. T. Y. Elizarenkova. Moscow, Nauka Publ., 1989. – Mandala X, hymn 15.
        12. The Laws of Manu (Manusmṛti) / transl. from Sanskrit. Moscow, Oriental Literature Publ., 1992.
        13. The Qur’an. Krachkovsky. Moscow, Nauka Publ., 1990.
        14. Assmann J. Death and Salvation in Ancient Egypt. – Ithaca: Cornell University Press, 2005.
        15. Davies D. J. Death, Ritual and Belief. – London: Bloomsbury, 2017.
        16. Civil Code of the Russian Federation (Part Three) of 26.11.2001 No 146-FZ.
        17. Federal Law of 27.07.2006 No 152-FZ “On Personal Data”.
        18. Bundesgerichtshof. Urteil vom 12.07.2018 – III ZR 183/17 // Neue Juristische Wochenschrift. – 2018.
        19. Revised Uniform Fiduciary Access to Digital Assets Act (2015) // Uniform Law Commission [Electronic resource].
        20. Facebook. Memorialized Accounts. Available at: https://www.facebook.com/help (accessed: 13.02.2026).
        21. Google. Inactive Account Manager. Available at: https://myaccount.google.com/inactive (accessed: 13.02.2026).
        22. Lastowka G. Virtual Justice: The New Laws of Online Worlds. – New Haven: Yale University Press, 2010.
        23. Boellstorff T. Coming of Age in Second Life: An Anthropologist Explores the Virtually Human. – Princeton: Princeton University Press, 2008.
        24. Edwards L., Harbinja E. Protecting Post-Mortem Privacy: Reconsidering the Privacy Interests of the Deceased in a Digital World // Cardozo Arts & Entertainment Law Journal. – 2013. – Vol. 32.
        25. Banta N. Inherit the Cloud: The Role of Private Contracts in Distributing Digital Assets // Fordham Law Review. – 2014. – Vol. 83.
        26. Meta Platforms, Inc.Social networking system that simulates a user using a language model : patent application US. – Publ. 2024. Available at: https://patents.google.com (accessed: 13.02.2026).
        27. Death isn’t the end: Meta patented an AI that lets you keep posting from beyond the grave [Electronic resource] // Business Insider. – 2026. Available at: https://www.businessinsider.com/meta-granted-patent-for-ai-llm-bot-dead-paused-accounts-2026-2 (accessed: 13.02.2026).
        28. Meta files patent for AI to post as users after death [Elektronnyi resurs] // Thairath. – 2026. Available at: https://en.thairath.co.th/money/tech_innovation/tech_companies/2913852 (accessed: 13.02.2026).
        29. Solum L. Legal Personhood for Artificial Intelligences // North Carolina Law Review. 1992. Vol. 70.
        30. Bayern S. The Implications of Modern Business-Entity Law for the Regulation of Autonomous Systems // Stanford Technology Law Review. 2016. Vol. 19.
        31. Hallevy G. Liability for Crimes Involving Artificial Intelligence Systems. Cham: Springer, 2015.
        32. Teubner G. Digital Personhood? The Status of Autonomous Software Agents in Private Law // Ancilla Iuris. 2018.
        33. Pagallo U. The Laws of Robots. Dordrecht: Springer, 2013.
        34. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence // Official Journal of the European Union. 2024. Art. 3.
        35. Kant I. Critique of Pure Reason. Moscow, Nauka Publ., 1998.
        36. Hegel G.V.F. Phenomenology of Spirit. Moscow, Nauka Publ., 2000.
        37. Dennett D. The Intentional Stance. — Cambridge, MA: MIT Press, 1987.
        38. Losev A.F. Dialectics of Myth. Moscow, Mysl Publ., 1991.
        39. Berdyaev N.A. O naznachenii cheloveka [On the appointment of a person]. Moscow, Respublika Publ., 1993.
        40. Gogol N.V. Dead Souls. Moscow, Khudozhestvennaya literatura Publ., 1984.
        41. Gogol N.V. Overcoat. sobr. soch.: in 14 vols. — T. 3.
        42. Gogol N.V. Nos // Poln. sobr. soch.: in 14 vols. — T. 3.
        43. Pelevin V.O. iPhuck 10. Moscow, Eksmo Publ., 2017.
        44. Pelevin V.O. Transhumanism Inc. — Moscow: Eksmo, 2021.
        45. Pelevin V.O. KGBT+. Moscow, Eksmo Publ., 2022.
  1. SCIENTIFIC PLATFORM “21 CENTURY” journal, # 1, April 2025, pp. 4-13, www.episs.ru; Scientific Platform “21 CENTURY” Journal, # 2, May 2025, pp. 92-126, www.episs.ru; Scientific Platform “21 CENTURY” journal, # 3, June 2025, pp. 59-69, www.episs.ru; Scientific Platform “21st Century” Journal, # 4, July 2025, pp. 24-56, www.episs.ru; Scientific Platform “21st Century” Journal, # 5, August 2025, pp. 3-42, www.episs.ru; SCIENTIFIC PLATFORM “21 CENTURY” journal, # 6, September 2025, p. 3-34, www.episs.ru; SCIENTIFIC PLATFORM “21 CENTURY” journal, # 7, October 2025, pp. 4-36, www.episs.ru. ↑
  2. Polybius. Universal History / transl. from ancient Greek. Moscow, Nauka Publ., 1986. (Description of the Imagines Maiorum, Book II. VI). ↑
  3. Li ji (The Book of Rites) / transl. from Chinese – Moscow: Oriental Literature, 1999. ↑
  4. The Bible. Books of the Holy Scriptures of the Old and New Testaments. Moscow: Russian Bible Society, 2011. ↑
  5. Ṛgveda / transl. from Sanskrit. T. Y. Elizarenkova. Moscow, Nauka Publ., 1989. – Mandala X, hymn 15. ↑
  6. The Laws of Manu (Manusmṛti) / transl. from Sanskrit. Moscow, Oriental Literature Publ., 1992. ↑
  7. The Qur’an. Krachkovsky. Moscow, Nauka Publ., 1990 ↑
  8. Assmann J. Death and Salvation in Ancient Egypt. – Ithaca: Cornell University Press, 2005. ↑
  9. Davies D. J. Death, Ritual and Belief. – London: Bloomsbury, 2017. ↑
  10. Civil Code of the Russian Federation (Part Three) of 26.11.2001 No 146-FZ. ↑
  11. Federal Law of 27.07.2006 No 152-FZ “On Personal Data”. ↑
  12. Bundesgerichtshof. Urteil vom 12.07.2018 – III ZR 183/17 // Neue Juristische Wochenschrift. – 2018. – S. 3178–3183. ↑
  13. Электронный ресурс: Revised Uniform Fiduciary Access to Digital Assets Act (2015) // Uniform Law Commission ↑
  14. Facebook. Memorialized Accounts. Available at: https://www.facebook.com/help (accessed: 13.02.2026). ↑
  15. Google. Inactive Account Manager. Available at: https://myaccount.google.com/inactive (accessed: 13.0.2026). ↑
  16. Lastowka G. Virtual Justice: The New Laws of Online Worlds. – New Haven: Yale University Press, 2010. ↑
  17. Boellstorff T. Coming of Age in Second Life: An Anthropologist Explores the Virtually Human. – Princeton: Princeton University Press, 2008. ↑
  18. Lastowka G. Virtual Justice: The New Laws of Online Worlds. – New Haven: Yale University Press, 2010. ↑
  19. Edwards L., Harbinja E. Protecting Post-Mortem Privacy: Reconsidering the Privacy Interests of the Deceased in a Digital World // Cardozo Arts & Entertainment Law Journal. – 2013. – Vol. 32. – P. 1–42. ↑
  20. Banta N. Inherit the Cloud: The Role of Private Contracts in Distributing Digital Assets // Fordham Law Review. – 2014. – Vol. 83. – P. 799–854. ↑
  21. Meta Platforms, Inc.Social networking system that simulates a user using a language model : patent application US. – Publ. 2024. Available at: https://patents.google.com (accessed: 13.02.2026). ↑
  22. Ibid. ↑
  23. Death isn’t the end: Meta patented an AI that lets you keep posting from beyond the grave [Electronic resource] // Business Insider. – 2026. Available at: https://www.businessinsider.com/meta-granted-patent-for-ai-llm-bot-dead-paused-accounts-2026-2 (accessed: 13.02.2026). ↑
  24. Meta files patent for AI to post as users after death [Electronic resource] // Thairath. – 2026. Available at: https://en.thairath.co.th/money/tech_innovation/tech_companies/2913852 (accessed: 13.02.2026). ↑
  25. Solum L. Legal Personhood for Artificial Intelligences // North Carolina Law Review. 1992. Vol. 70. P. 1235. ↑
  26. Bayern S. The Implications of Modern Business-Entity Law for the Regulation of Autonomous Systems // Stanford Technology Law Review. 2016. Vol. 19. P. 104. ↑
  27. Hallevy G. Liability for Crimes Involving Artificial Intelligence Systems. Cham: Springer, 2015. P. 102. ↑
  28. Teubner G. Digital Personhood? The Status of Autonomous Software Agents in Private Law // Ancilla Iuris. 2018. P. 35–78. ↑
  29. Pagallo U. The Laws of Robots. Dordrecht: Springer, 2013. P. 101–103. ↑
  30. For more details, see the author’s works: Journal SCIENTIFIC PLATFORM “21 CENTURY”, # 1, April 2025, pp. 4-13, www.episs.ru; Scientific Platform “21 CENTURY” journal, # 6, September 2025, pp. 3-34, www.episs.ru; Scientific Platform “21 CENTURY” journal, # 7, October 2025, pp. 4-36, www.episs.ru. ↑
  31. Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence // Official Journal of the European Union. 2024. Art. 3. ↑
  32. Kant I. Critique of Pure Reason. Moscow, Nauka Publ., 1998. — P. 131. ↑
  33. Hegel G.V.F. Phenomenology of Spirit. Moscow, Nauka Publ., 2000. — P. 111. ↑
  34. Dennett D. The Intentional Stance. — Cambridge, MA: MIT Press, 1987. — P. 17. ↑
  35. Losev A.F. Dialectics of Myth. Moscow, Mysl Publ., 1991. — P. 25. ↑
  36. Berdyaev N.A. O naznachenii cheloveka [On the appointment of a person]. Moscow, Respublika Publ., 1993. — P. 62. ↑
  37. Gogol N.V. Dead Souls. Moscow, Khudozhestvennaya literatura Publ., 1984. — S. 67–69. ↑
  38. Gogol N.V. Overcoat. sobr. soch.: in 14 vols. — T. 3. — S. 145–176. ↑
  39. Gogol N.V. Nos // Poln. sobr. soch.: in 14 vols. — T. 3. — S. 57–75. ↑
  40. Pelevin V.O. iPhuck 10. Moscow, Eksmo Publ., 2017. ↑
  41. Pelevin V.O. Transhumanism Inc. — Moscow: Eksmo, 2021. ↑
  42. Pelevin V.O. KGBT+. Moscow, Eksmo Publ., 2022. ↑