1. Introduction

Can administrative law regulate the artificial intelligence? This question has been asked frequently in recent years. The purpose of this contribution is not to evaluate how administrative law supports or, conversely, limits the development of information technologies, or to interpret existing legislation. Attention is focused on how public administration and administrative law are changing in the digital environment.

2. Technological challenges

As Jean-Bernard Auby explained, the digitalization of public administration is not the result of a revolutionary political change affecting public administration and administrative law. It is a mutation that acts gradually and often unexpectedly, as the great technological revolutions often do^[1].

It is argued that regulation is slower than technological development. But we should not oversimplify this stereotype. In her insightful article,^[2] Lyria Bennett Moses criticizes lawyers for jumping in quickly when something is new and exciting, while fading away when the issues become more mundane but no less important. The challenge of regulatory connectivity, however, is constant. Technologies will not always be new, but they will always change and present new legal and regulatory challenges. Because change is constantly occurring, we need to think broadly about how to minimize risks of harm and protect important values in an evolving sociotechnical environment, rather than simply asking how technology should be regulated. To the extent that some technologies create new environments (such as cyberspace and virtual worlds), these issues primarily concern information and communication technologies.

Technology is a very broad concept. It can refer to the practical use of knowledge, the way in which tasks are accomplished using technical processes and methods, or the construction of a certain activity. Is technology also governance? This question evokes the philosophical legacy of Michel Foucault. Let us recall his words: «Government is possible only when the strength of the state is known: It is by this knowledge that it can be sustained. The state’s capacity and the means to enlarge it must be known … A government, therefore, entails more than just implementing general principles of reason, wisdom, and prudence. A certain specific knowledge is necessary: concrete, precise, and measured knowledge as to the state’s strength»^[3]. Public administration in this sense is “art of government” that combines institutional, discursive, and practical techniques.

As Villadsen’s book amply demonstrates, Foucault’s legacy provides a toolkit suitable for analysing developments that have rapidly spread and are associated with surveillance capitalism and artificial intelligence^[4]. It is also evident that the current reassessment of Foucault’s findings therefore encourages a change of focus – from revealing the visible mechanisms of power to examining the subtle architectures that organize behaviour, perception, and legibility^[5].

Foucault explained that the theory of the legitimate basis of sovereignty cannot be relied upon as a means of describing the ways in which power is exercised under such sovereignty. Power operates in a network, and within that network, individuals can constantly carry and exercise that power. In other words, the theory of the legitimate basis of sovereignty cannot be relied upon as a means of describing the ways in which power is exercised under such sovereignty^[6].

By the words of E. Ongaro, philosophy of public administration can enable a better connection between the field of philosophy of information and the field of e-government and digital governance^[7]. The philosophical question about nature and meaning of information is something we are asking ourselves today and has nothing to do with futurological thinking. As Luciano Floridi states: «Are we really going to be wearing some kind of VR headsets to live in a virtual or augmented world created and inhabited by AI systems? Consider how many people today are reluctant to wear glasses even when they seriously need them, just for aesthetic reasons. And then, are there feasible AI solutions in everyday life? Are the necessary skills, datasets, infrastructure, and business models in place to make an AI application successful? Futurologists find these questions boring… It is the bad diet of junk fast food for thoughts and the curse of the airport bestseller. We must resist oversimplification»^[8].

It is necessary to step back a bit and look at the overall trend of artificial intelligence development. We live in a world of narrow artificial intelligence. However, our reflections on law and public administration go further. They are set against the backdrop of a current moment that arguably represents a historic nadir in the adoption of AI and public trust in many sensitive areas. Yet this current scepticism appears increasingly fleeting. All indications are that future generations will increasingly accept, and even expect, the integration of AI into sensitive tasks, including those traditionally reserved for human discretion. This anticipated pervasive integration heralds a cultural shift in trust, which in turn implies a widespread willingness to embrace these technologies even within the most sacred of government processes.

Information is generally understood as the basic unit of knowledge. The primary information technology of governance was writing. Evidence can be found in the work of archaeologists and historians. Let us give an example. Miroslav Bárta reminded us that today it is difficult to imagine a turning point in civilization when the ancient Egyptians could start using the first simple signs and assigning them meaning for the purpose of preserving and formally transmitting information. However, it took several centuries before sentences with abstract meanings could be recorded in writing^[9]. When this happened in the second half of the 3rd millennium BC, a very early form of administrative law also emerged.

«Justice is beautiful and its excellence is permanent, unchanged from the time immemorial of the beginnings of kingdoms, and although he who disregards its rules is punished, it is still far from the uneducated man»^[10]. We find this sentence in the famous text The Maxims of Ptahhotep. When a conference on the principles of good governance was held in the Czech Republic in 2006, Judge of the Constitutional Court Stanislav Balík surprised some participants by claiming that all the principles currently being discussed, except for speed and transparency, were well known to the high-ranking Old Kingdom official Ptahhotep^[11].

According to Judge Balík, it is impossible to determine exactly when Ptahhotep’s teachings fell into oblivion in the distant past. What is important is that the teachings were recorded and that they were rediscovered. Ptahhotep urges officials not to boast about their knowledge and to express themselves only when they are absolutely certain. Only those who know what they can really solve should propose a solution. This is a remarkably current interpretation that goes beyond the ethical dimension and emphasizes that data and nonsense are abundant, while knowledge is scarce. Already in the Bronze Age, governance was understood and implemented as a technology, albeit not permanently. Ancient Egyptian law was associated with the power of the pharaoh, who, as a god on earth, issued decrees enforced by officials, but at the same time it was based on the principle of divine order, truth and justice.

This reflects Foucault’s observation that what is truly important for our modernity is not so much a nebulous form of mastering society as it is the governmentalization of the state. Information technology used for governance has not changed for centuries and has relied on human knowledge and empirical data. Development has not been continuous. Coercion has usually been a more effective technology than knowledge. This is clear in the study of a phenomenon that can be somewhat uneducatedly called proto-administrative law which refers to early legal norms and practices governing public authority and officials before the emergence of modern administrative law as a distinct field. History of governmentality in this sense is the process, through which the society of laws transformed into the administrative state gradually becomes governmentalized^[12].

In the second half of the last century, the philosophy of public administration raised the urgent question of whether anthropogenic information capacity has been exhausted. As E. Ongaro concluded, the impact of artificial intelligence on public administration and the dramatic changes in interactions between the state and citizens caused by revolutionary technological innovations encourage the use of the intellectual power of philosophy, which deals with questions that cannot be solved by standard scientific methods^[13].

3. Virtual praxis

Information permeates and shapes reality. If information technology changes our experience of reality, it moves us towards simulated worlds. Such a shift may become a preferred approach in public administration, for example, due to the priority given to fast response and user comfort. We are willing to accept public administration as it is, but also as it is not. The automation of decision-making often seems only to reflect the future.

Digitalization has opened a wide space for the study of virtual properties, which evokes a mirage. As an example, in the first decade of this century, Sweden and Estonia temporarily opened their virtual embassies using the video game Second Life. Other states were inspired by this and began to provide some consular services to potential applicants in this way. A mirage, like an image in a mirror, is something that appears and manifests itself as a real phenomenon.

Immersive technologies blur the distinction between the real and virtual worlds. Sure, we can usually file a complaint or official request in thirty seconds from our tablet without wanting to see the office reflected. But aren’t we missing a better experience? If not, we are not interested in immersive technologies because we don’t really want to reach a state of consciousness where the environment in which we communicate can be subconsciously considered real.

Generative AI poses a particular problem because of its ability to approximate the text and its tendency to confabulate. A more conservative prediction of AI’s ability in the medium term is that it will be useful in circumstances where uncontroversial categorization is possible, such as where there is a clearly established legal rule that can be applied to undisputed facts. It is not yet clear whether an expert system will really help in cases where facts and rules are less easily discernible, or where other categorization problems arise^[14].

At the turn of the millennium, American administrative scholar and former advisor to the Reagan administration, James Colvard, published a reflection on the fact that the main manifestations of the depersonalizing and non-regulatory trend in public administration are not primarily associated with information and communication technologies, but with managerial innovations that depersonalize the processes of governance within the bureaucracy. He compared the situation in the administration to what is happening in sports clubs. While sports clubs used to cultivate young players on their farms and fans spent several years getting to know them, worrying about their failures and rejoicing in their successes, now players are simply bought. Fans therefore need a reference manual on their mobile phones to be able to recognize players and are much less tolerant of failure, because they do not see players as individuals, but rather as luxury goods. According to Colvard, the problems of government are a kind of caricature of the objectivist concept of management, which assumes that all actions can be quantified, and thus everything becomes objective. Technology rather complicates the situation. Today, it is possible to communicate with an organization and never speak to a live person. Even when interacting with a person, people mostly deal with impersonal messages on modern machines. The sense of individual responsibility and accountability is lost in such an environment. Future managers must learn to use modern technologies while also approaching them in a humane manner^[15].

Contemporary public administration can function as a virtual organization that uses digital technology and artificial intelligence to achieve its goals and fulfil its tasks, rather than physical presence and personal contact. In cyberspace, large volumes of data can be processed, decisions can be made automatically, and intelligent behaviour can be imitated. However, it is not just about demanding activities that would be impossible without digital interaction. We can hardly imagine that offices could function without computers even in routine agendas. Even in everyday practice, we find ourselves in the space of augmented reality connecting the physical world with the world of its digital twin. Relevant information is located somewhere, but it is also virtually present at any accessible point of the network. Therefore, we can also define virtuality as a property of phenomena that manifest themselves as if they were real.

Virtuality is inherently linked to the expansion of information and communication technologies. If we also understand public administration as a technology of governance in cyberspace, we can choose between an optimistic and a pessimistic vision. Advanced information and communication technologies can either help or harm. The reality mediated by “smart” machines is an image of artificial administration, which can lead to de-bureaucratization and a reduction in the complexity of public administration. However, new technologies can also have an oppressive potential and lead to public administration becoming an unbearable control monster.

The term virtualization, which has been used since antiquity, refers to a technologically determined change in the formal aspects of a phenomenon while preserving its essence. The purpose of virtualization is to get rid of the problems associated with the old form, while the problems associated with the new form are supposed to be less severe.

The principle of virtuality deals with regulation in an environment where human action is mediated by technology. Virtuality is gaining strength given that cyberspace has become a social arena, which is entered by all public administration actors who use advanced information technologies to interact. In cyberspace, public administration communicates, makes decisions and controls as in reality. Decisions that have a significant impact on an individual or a community should be acceptable and satisfactory, also regarding the required level of autonomy of artificial interaction.

When social relations are mediated by technologies based on artificial intelligence, not only credibility is at stake, but also flexibility, allowing adaptation to different situations. Supporting active human participation in conditions of artificial communication and automated decision-making is also key.

Virtualization is associated with concerns about losing control over key decisions. This point of view is expressed in the discussion of the right to human decision-making, or the right to well-calibrated automated decision-making. Complex technology raises concerns about the ability to understand artificial reasoning and its methods of classification, personalization and prediction. Therefore, there is talk of the problem of readability of artificial decision-making and the expanded scope of artificial intelligence regulation. This intention guides the legal regulation, both in terms of the right to digital services and in relation to the algorithmic rights and information security. The right to use a digital service corresponds to the obligation to provide this service. However, alternatives should always be offered that protect the rights of those persons who decide not to use digital technology. A digital act by which a clearly identifiable person clearly expresses his will is, in terms of legal effects, equivalent to a legal act performed in another form. The expression of will made must be capable of being retrospectively identified by the public administration information system.

4. Continuity and change

Continuity is one of the attributes of public administration. Nowadays, we often encounter the view that the only constant that can be found in public administration and, by extension, in administrative law, is change. If this is the case, continuity manifests itself in continuous change.

In the performance of public administration, the continuity focused on fulfilling future goals is manifested in the fact that the closer the goal is, the more difficult it is to achieve this goal. The reasons are obvious: various winding paths lead to goals and decision-making is marked by external and internal influences. In this context, A. Roberts does not hesitate to recall a thesis known from the theories of multi-level governance: governments that ignore systemic pressures will not survive any more than companies that ignore market signals. Leaders must therefore learn to make decisions in an environment where key information is missing and where corrupt officials lurk at every turn^[16].

Public administration is sensitive to the environment and circumstances. It balances between lability and rigidity, so that it can sometimes seem like a ship that is being kept in one place in a stormy sea. Administrators seek the optimal degree of variability in decision-making in public administration. Law and public interest are correctives for the performance of public administration in a continuum of change. While public administration is a continuous purposeful activity that, in a complex environment, seeks a balance between what was and what will be, the judiciary can be considered a relatively static activity in which an independent judge finds the law that applies here and now.

The judicial finding of law hic et nunc tells whether the balance in public administration is currently being found, provided that administrative law is sufficiently clear. H.L.A. Hart wrote in the middle of the last century about two handicaps that get us into trouble whenever we try to clearly regulate behaviour using general norms. The first handicap is our relative ignorance of the facts, the second is our relative vagueness of purpose^[17]. Public administration is continuously coping with these handicaps, but at the cost of administrative law becoming increasingly specialized, fragmented and pluralistic. At the same time, this means that the level of complexity of legal decision-making is increasing.

We wonder about it will be easier to negotiate with a robot than with a living official and whether bureaucracy can be replaced by infocracy. The use of artificial intelligence tools leads to the question of whether the standards of technologists will surpass those of lawyers and whether the current legal principles in public administration can be sufficient.

The development of legal technologies enabling machine analysis of law and virtualization has accelerated in the last two decades. At the beginning of this wave, Volker Boehme-Nessler noted that what is possible in cyberspace has not yet been defined by democratically legitimized legislators, but by technical standards and software. Engineers and programmers have thus become legislators^[18].

R. Susskind had no doubt a few years later that most of the systems that will replace us in the legal profession will not work like us. Impact analysis, which often relies on an outdated distinction between routine and non-routine work, significantly underestimates the extent to which artificial intelligence will take over human work^[19]. In his latest book, the author reminds us that the defining global human endeavour of the second quarter of the twenty-first century is to properly integrate artificial intelligence into the activities in which smart machines are used. In doing so, it is better to be guided by Plato, Aristotle, and Kant rather than – with respect – Sam Altman, Elon Musk, and Mark Zuckerberg. The administration of justice is no exception, but there are many doubts about whether artificial intelligence is being “harnessed well”. When it comes to regulation, Susskind is relentless. For more than fifty years, we have complained that legislation and court decisions lag behind new technologies by many years. The explosive development of artificial intelligence is making the situation even worse. It could even lead to artificial intelligence starting to draft laws in natural language and executable code^[20].

In the last decade, the development of information technology has accelerated so much that we are discussing the prospect of artificial public administration now when the current narrow artificial intelligence will be overcome. The development of artificial intelligence will probably slow down due to economic and energy demands. However, machine data processing is already generally intelligent today, albeit fundamentally different from human knowledge.

We can identify with the considerations that Cary Coglianese has made^[21]. Advances in explainable artificial intelligence techniques likely will make automation still more compatible with long-standing administrative law values. It is perhaps obvious that automation can contribute to the goal of expert administration. Automation could very well drive the demand for still greater specification and clarity in statutes. Automated government might be smarter, more democratically accountable, and even more fair. But it could also lack feeling, even more than sterile bureaucratic processes do today. The empathy demands that administrative agencies provide opportunities for human interaction and for listening and expressions of concern. If administrative law in an automated state is to adopt any new rights, society might be better served if courts avoid the recognition of a right to a human decision and consider and seek to define a right to human empathy. Let us just add that Ptahhotep already wrote: «It is more valuable for a supplicant to be heard than to have his request granted»^[22].

5. Conclusions

The term virtual administrative law is not commonly used. A very close and more frequently used term is digital administrative law as a concept expressing that new information technology has become an integral part of public administration. When considering digital administrative law, lawyers usually combine the principles of general administrative law with technological phenomena such as narrow artificial intelligence, big data, and digital platforms.

Narrow artificial intelligence is referred to as lacking the ability to generalize or understand context beyond predefined functions. But it depends on how we define general intelligence. If general intelligence required the ability to quickly and accurately perform all cognitive tasks, there would probably be no human being who could do it. People have general intelligence, but to varying degrees. Human brain learns from data and looks for certain rules in it. It doesn’t matter how we get to the result, but what the result is. The assumption that artificial intelligence autonomously overcomes its current scope towards more advanced forms of cognitive abilities also puts legal reasoning in a new light. Artificially created legal sentences can be generated in such a way that they look plausible. However, looking plausible is something completely different from corresponding to natural reality.

Considerations about public digital law are usually closely related to the question of how public institutions should govern new technology. Virtual administrative law fills a complementary space that addresses how technology transforms administrative practice (public administration with artificial intelligence) and whether this change can be accepted as knowledge (public administration through artificial intelligence). Illustratively speaking, virtual administrative law is a concept that reveals the physiognomy of public administration in which digital rules are applied.

As Pierre Lévy explained, virtuality does not refer to some false or imaginary world. Virtualization is the very dynamic of a shared world^[23]. The virtualization of legal relationships as a way of sharing reality may have been lightly hoped to eliminate unnecessary formalism. However, both law and technology have become more complex and difficult to communicate than they were at the dawn of the digital age. However, there is little doubt about anthropocentric governance. Humans interact with anthropomorphized artificial intelligence. Even though people are aware that they are interacting with artificial systems, they attribute human-like qualities to these entities.

It remains only to confirm that if administrative agendas are digitalized, interfaces become simultaneously virtual and difficult to alter^[24]. It is evident that technologies mediating interaction between authorities and individuals will influence the perception of administrative-legal relations.

It is an open question where the limits of depersonalization extend. Virtuality is asserting itself as a new paradigm in the conditions of digitalization. The opinion that virtualization marks the twilight of anthropogenic public administration is not exceptional. It is an open question how far depersonalization in public administration can go. If virtualization sounds like the swan song of anthropogenic public administration, human centrality is remembered more than it is lived. However, if artificial public administration remains under our control, virtualization has the nature of a meta-principle in the sense that the principles of general administrative law are interpreted with respect to the way in which new information technologies change the exercise of public power.

1. See J.-B. Auby, Administrative Law Facing Digital Challenges, in European Review of Digital Administration & Law, 1-2, 2020, p. 15. ↑
2. L.B. Moses, How to Think about Law, Regulation and Technology: Problems with “Technology” as a Regulatory Target, in Law, Innovation and Technology, 1, 2013, p.11-18. ↑
3. M. Foucault, The political technology of individuals, in L. H. Martin, H. Gutman, P. H. Hutton (eds.), Technologies of the Self: a seminar with Michel Foucault, University of Massachusetts Press, Amherst, 1988, p. 151. ↑
4. See K. Villadsen, Foucault’s Technologies: Another Way of Cutting Reality, Oxford University Press, Oxford, 2024. ↑
5. See H. Sahakyan, A. Gevorgyan, A. Malkjyan, From Disciplinary Societies to Algorithmic Control: Rethinking Foucault’s Human Subject in the Digital Age, in Philosophies 4, 2025, p. 73. ↑
6. See C. Gordon, Governmental rationality: an introduction, in G. Burchell, C. Gordon, P. Miller (eds.), The Foucault effect: studies in governmentality, University of Chicago Press, Chicago, 1991, p. 7. ↑
7. E. Ongaro, Connecting Philosophy and Public Administration, Palgrave Macmillan, Cham, 2025, p. 126. ↑
8. L. Floridi, The Ethics of Artificial Intelligence, Oxford University Press, Oxford, 2023, pp. 52-53. ↑
9. See M. Bárta, Příběh civilizace, Academia, Praha 2016, p. 146-148. ↑
10. Ptahhotep, Naučení vezíra Ptahhotepa, pp. 83-86, in B. Vachala (ed.), Moudrost starého Egypta, KPK, Praha 1992, p. 13. ↑
11. S. Balík, Maximy vezíra Ptahotepa – starověký etický kodex úředníka, in D. Hrabcová (ed.), Principy dobré správy, Masarykova univerzita, Brno, 2006, pp. 21-26. ↑
12. See M. Foucault, Governmentality, in G. Burchell, C. Gordon, P. Miller (eds.), The Foucault effect: studies in governmentality, University of Chicago Press, Chicago, 1991, p. 103. ↑
13. See E. Ongaro, Connecting Philosophy and Public Administration, Palgrave Macmillan, Cham, 2025, p. 33. ↑
14. See Lord Hodge, The Law and AI: where are we going? Lecture at De Montfort University, Leicester, 30 November 2023, available on www.supremecourt.uk. ↑
15. See J. Colvard, Restore the Human Touch, in Government Executive, 1, 2000, available on www.govexec.com. ↑
16. See A. Roberts, The Aims of Public Administration: Reviving the Classical View, in Perspectives on Public Management and Governance, 1, 2018, p. 76. ↑
17. See H.L.A. Hart, The Concept of Law, Clarendon Press, Oxford, 1961, p. 128. ↑
18. See V. Boehme-Nessler, Unscharfes Recht Überlegungen zur Relativierung des Rechts in der digitalisierten Welt, Duncker & Humblot, Berlin, 2008, p. 202. ↑
19. See R. Susskind, Online Courts and the Future of Justice, Oxford University Press, Oxford, 2019, p. 51. ↑
20. See R. Susskind, How to Think About AI: A Guide for the Perplexed, Oxford University Press, Oxford 2025, pp. 107 ff. ↑
21. See C. Coglianese, Administrative Law in the Automated State, in Daedalus, 3, 2021, pp. 108-116. ↑
22. Ptahhotep, Naučení vezíra Ptahhotepa, op. cit., p. 16. ↑
23. See P. Lévy, Becoming virtual – reality in the Digital Age, Plenum Trade, New York, 1998, p. 184. ↑
24. See J. Raso, Implementing Digitalization in an Administrative Justice Context, in M. Hertogh, R. Kirckham, R. Thomas, J. Tomlinson (eds.), The Oxford Handbook of Administrative Justice, Oxford University Press, Oxford, 2022, p. 526; S. Ranchordás, The Invisible Citizen in the Digital State: Administrative Law Meets Digital Constitutionalism, in C. van Oirsouw, J. de Poorter, I. Leijten, G. van der Schyff, M. Stremler, M. De Visser (eds.), European yearbook of constitutional law 2023: Constitutional law in the digital era, T.M.C. Asser Press, The Hague, 2024, p. 23. ↑