Research Article | | Peer-Reviewed

Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization

Received: 2 April 2025     Accepted: 14 April 2025     Published: 29 April 2025
Views:       Downloads:
Abstract

The power’s digitalization in the era of artificial intelligence refers to the deep coupling of technical logic and power logic after digital technology is embedded in the power operation mechanism, thereby promoting a more comprehensive technological transfer in power distribution. With the assistance of technology, power has moved behind the scenes, but the power's ability to control society has been enhanced. The external characteristics of this phenomenon are strong technicality, concealment and intrusion. There are three aspects of the internal logic that led to its emergence: human digital survival has become the norm, the diffusion effect of power is significantly better than its coercive power, and the black box operation space brought by digital technology and people's dependence on digital codes. Using comparative research, empirical research and literature analysis, the challenges encountered by traditional administrative rule of law are explained. The basic principles of modern administrative rule of law established since the 19th century are quietly changing, especially in two aspects: digital technology has eliminated the basic rights of citizens represented by privacy rights and basic political rights, making discretionary power difficult to control and the definition of traditional legal responsibilities difficult. These are challenges to the basic principles of modern administrative law since the 19th century. Controlling or narrowing the legal gaps caused by the digitization of power, combining public and private law protection of citizens' right to information, and establishing a legality review system for technology operations centred on the controllability of digital technology and public participation should become a feasible approach.

Published in Journal of Public Policy and Administration (Volume 9, Issue 2)
DOI 10.11648/j.jppa.20250902.14
Page(s) 91-99
Creative Commons

This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.

Copyright

Copyright © The Author(s), 2025. Published by Science Publishing Group

Keywords

Digitalization, Legal Risks, Neutrality, Traditional Administration, Comparative Methodology

1. Introduction
The transformation of power from a natural state to modernization has always been accompanied by the emergence and iteration of technology, from the agricultural age to the industrial age and then to the digital age, power and technology are closely integrated. From this perspective, the power’s technologization is a dynamic and ongoing process, the degree of close integration between power and technology is positively correlated with the profound process of technology as a tool in changing human life.
In the era of artificial intelligence (AI), a set of governance technologies that achieve the purpose of power through the integration and interaction of power and technology, as tools, is called digital governance or digital government. So far, almost all research has been centered around digital and AI's assistance to power and has talked about the new changes in power, the correlation between algorithms and power has become a fashionable topic. When power increasingly actively and enthusiastically embraces digital technology (DT) , the changes in power boundaries and order would be disrupted by algorithms in a covert way. Digital power brings not only efficient social governance, but also a considerable number of legal risks.
Through embedding DT into the operation scenario of power, deeply integrating technical work and power logic, further promoting the operation and distribution of national powers, especially administrative power, more widely, deeply and technically processed, during this series of processes and operations, national powers enter the behind the scenes , its ability to control society and its order is enhanced, “rapid expansion of power” is achieved, and at the same time, “power concealment” is also achieved. How to prevent the situation where “power arbitrariness” occurs under the guise of technological innovation. All thoughts and efforts need to have a clear PD’s definition and provide legal risks and some coping plans caused by the digitization.
2. PD’s External Characteristics
2.1. Technicality
DT is a complex system, such as information, algorithms, and code. When it becomes a medium or channel for the operation of power, it usually manifests itself as a technical requirement for the object or business that holds power. A typical example is criminal investigation using big data, including a series of complex processes, such as data collection, classification and cleaning, data mining and comparison , the requirements for complex procedures and professional skills in traditional criminal investigation generally will be reduced, oral questioning, on-site record keeping and other face-to-face activities are still the main scenes of traditional investigation activities.
The investigative behavior’s essence is the functional expression of police power as one of the national administrative powers. When big data investigation becomes an important medium for the exercise of police power, it is not difficult to find that DT has reshaped the specific investigative behaviors. In a considerable number of cases, before the traditional investigation begins, the police obtained clues through data analysis, this has indeed improved the efficiency of traditional criminal investigations. However, the development of technology and the combination of DT and police power will inevitably put forward higher requirements on the professionalism of administrative personnel .
2.2. Concealment
During the close combination of power and technology, power technologization is inevitable, DT has become a manifestation, while power is hidden behind the scenes, it frequently occurs in the governance of public affairs. In the Covid 19 that swept the world, the Chinese government had fully implemented the “health code” model to prevent and control the epidemic and prevent the widespread spread of the virus. Specifically, everyone used their mobile phone number to bind the “health code”, and the health code was used to detect the user's activity range, time and other factors. Through DT, each person was given a unique set of digital codes or digital identities , and the “green code pass” was used as the basic rule to participate in social mobility. Within the scope of the health code, the user's travel time, area, contacted people and other information were all controlled and retrieved by the local government, mainly the health administrative department.
From the perspective of governance effect, the “health code” as a DT’s application indeed achieved significant governance by introducing it into emergency management. However, from PD’s perspective, administrative power was reasonably hidden behind the scenes and was no longer transparent as advocated in traditional administrative rule of law. The difference and distinction between secrecy and transparency is precisely the logical divergence in the operation of administrative power under traditional administrative rule of law and AI. From DT to the application of health codes, “people's right to judge and choose is excluded” and “whether it is questioning or resistance, obedience is the primary obligation” .
2.3. Intrusiveness
PD has triggered a deeper intrusion into people's lives, with the most significant being comprehensive control of administrative power over people's lives. Taking big data investigation as an example, it does have the advantages of foresight and initiative, but this means that the mastery and in-depth massive information through big data technology is suspected and possible to interfere with the rights of personal information and privacy, investigations assisted by DT fully demonstrate that control over daily life has far exceeded the time and space required to define criminal violations.
DT and human digitization (HD) have become important prerequisites for the effectiveness and scope of power. On the one hand, power has penetrated more and more comprehensively into daily life. Power can be expressed in a variety of forms, such as platforms, apps, social media, etc., and can influence or intervene in a certain individual at anytime and anywhere, and even such influence and intervention may be arbitrary. On the other hand, as we enjoy the AI’s advantages, human transparency has become unavoidable, power has achieved the observation and monitoring of human bodies, health, and activities. This kind of monitoring is even worse than the exposure of privacy rights, “human bodies are breached” and “human health is fully displayed” .
3. PD’s Inherent Logic
3.1. Human Digital Life Has Become the Social Scenes
The digitalization of people's lifestyle is mainly reflected in instant messaging, information acquisition, online office, business transactions, online entertainment and online public services. Massive amounts of data are generated and collected every day, and society is the main field for data production, distribution and exchange . According to incomplete statistics from the Chinese government, from 2017 to 2023, China's data output increased from 2.3 ZB to 6.6 ZB . The processing and analysis of these data can achieve detailed analysis and in-depth observation. Considering the necessary exchange of convenience in life, DT is assumed to have special qualifications for data processing , and people with DT can restore a real individual in the algorithm program through detailed analysis of data. After this in-depth analysis and algorithm description, the complexity of people is diluted and becomes an object that can be analyzed and quantified .
Thus, in the digital society, people finally become a transparent one. After being made transparent, people’s behaviors and thoughts are dominated and constrained by DT. What’s worse, with the “strong technology dependency syndrome” , people unconsciously develop a default in the logic of technology. This default structure is that people obey the control and shaping of individual bodies and minds and apply the logic of obedience to various disciplinary scenarios. Technology logic amplifies the appearance of technology’s control over people but hides the essence of power discipline.
3.2. Power Diffusion Is More Effective
Michael Mann divided power into authoritative power and diffused power according to the different ways of exercise, the former emphasized command and obedience, while the latter emphasized understanding and cooperation. National police power is a typical administrative power with the above two characteristics. To control the disciplinary mechanism, the exercise of power by the police system is supported by force and embodies violence. Police power is a typical authoritative power, which controls the public through command and obedience , this is a scene that occurs in the traditional administrative rule of law. In the AI’s era, public understanding obtained by big data investigation or big data police law enforcement is “natural, moral or derived from self-evident common interests” . Through PD, police power has diffusion.
Take the Tian Wang system as an example, it can densely cover a wide geographical area and achieve full surveillance of the geographical space. From the public's perspective, ordinary people have shown great tolerance and have not even questioned their concerns about privacy ethics and technical ethics. What's more, they are still tolerant of the public security organs’ violations of citizens’ privacy rights by means of monitoring and positioning to maintain social stability . The reason is that the public recognizes and accepts the diffuses of police power, even in a panoramic open surveillance state, ordinary people think this is a kind of security care . This common mentality has promoted the continuous integration and influence of police power, police agencies and even individual police officers are increasingly finding that diffuse power encounters less resistance , and the traditional command-and-obedience model of police power might bring about vicious social incidents.
3.3. Black Box Operation Becomes Easy
While digital governance is advocated, administrative power is expressed follows the process of digital reform. When “improving the level of digital government construction” becomes a task, the universal application of DT can realize “the operation mode, business process and service model” , and “build a digital government decision mechanism” . Under such a policy background and operation logic, the logic of automated and intelligent administration conforms to the technical logic of “input-output” , the traditional administrative activity mode is switched to the technical mode of “input-output”.
One problem cannot be avoided. The foundation of automated administration, from a structural level analysis, should at least include three parts: input layer-hidden layer-output layer . What is regarded as the input layer is the administrative application, which is the start-up procedure in the traditional administrative procedure, and what is regarded as the output layer is the administrative behavior or administrative activity, which is the legal relationship that creates a legal connection between administrative power and administrative counterparts. The administrative procedures and their operation in the middle are regarded as the hidden layer. This hidden layer cannot be participated in under the operation of technologies by PD, and there is suspicion of entering a black box .
The most representative one is administrative discretion. Administrative discretion is the controversial part of the rationality requirements of administrative power. How administrative discretion implements administrative power in the middle of “input-output” is inconsistent with the openness and transparency of administrative procedures and hearing procedures advocated in traditional administration . The logic of dark box operation has achieved the effect of invisibility in PD’s process.
3.4. Code Control Becomes Dependent
Digital life controlled by code causes people to have a “strong technical dependence” . Under the dual effects of strong technical dependence and virtual reality, people are becoming more digitalized, and the logic of PD's regulation and control of people has changed, the regulation or management of people can be divided into “regulation of code” and “control by using code rules” . Specifically, “regulation of code” is manifested as the digital identity constructed based on people's life and behavior data becoming an individual who finds a way to pass and entre in the digital society, making the recognition of code in different situations even exceed the real connection with the subject.
As it explained earlier, the “health pass code” promoted by the Chinese government, people had more trust than people have in each other. As a part of social group life, it is particularly important when two or more parties are unfamiliar. On the other hand, power controls codes through DT, further realizing barrier-free, all-time and space regulation of people, using the generation of code rules to deeply integrate virtual and real, and realize the interaction of digital virtual space. As various platforms and mobile Apps that are important media, this is the “Memorandum of Mutual Understanding” unilaterally generated by the designer, as a user, there is no right and qualification to formulate, modify rules, etc.
Compared with the relief and correction advocated in traditional legal procedures, the constraints-in-advanced of the code do achieve “full prevention before people implement the behavior” , and even “deprive individuals of the ability to choose whether the rules are violated” . Under the blessing of DT, facing power, the common person has lost the ability to resist and choose.
4. PD’s Legal Risks and Analysis
When DT has been embedded in the power operation procedures, it promoted standardization, legitimization, and proceduralizing of power. However, PD has brought about hidden and invasive negative effects. Some of these negative effects began to become apparent in people’s daily lives, while others have been concealed by technological neutrality.
4.1. PD Violates Citizens’ Basic Rights
Beyond the physical space, the digital virtual space has become a space for public activity, where the public's behaviors are preserved in data and code, and when the data is large enough, individual cannot hide himself in the DT’s front, and ultimately “submits” to the digital control, and the citizen's right to privacy is gradually eliminated. Individuals in the digital life model “have become accustomed to using their personal informational images to refer to real individuals” and construct their digital identities through code or data, and even though anonymity systems guarantee that each individual would not be nakedly displayed on the Internet in a certain procedure, but, data owners can still use their data to discern citizens’ identity privacy, information privacy, psychological privacy and its correlation analysis, “from one or two kinds of privacy information species deduce other information and data” , the digitalization of people will lead to the transparency of people. The transparency of people dilutes the material existence of people in the spiritual existence in front of the data, people become unreserved and difficult to escape, DT is embedded in the operation of power, the individual right to privacy is limited or even extinguished, and everyone becomes a producer of data, but not a holder of data.
In the public sphere, “the government is the largest owner of citizens’ information, and it possesses not only personal data, but also the government is the powder keg of citizens’ core privacy” . Based on public interest and management needs, the government can reasonably and justifiably access and use citizens’ information, applying big data technology in investigative activities, embedding automated procedures into administrative enforcement processes, embedding AI into judicial processes, etc. Behind all these phenomena is the foundation of the power's possession of huge quantities of citizens’ data, and the power will enjoy the data properly. “The rationality of technology is transformed into the rationality of politics” , and “the use of technology seems to gain greater public inclusion” .
In the “digital panorama of open surveillance” , power has the opportunity to hide behind DT, to spy on everyone without any dead angle, and even to “pry into the psychological privacy of the individual” , which breaks through the barrier of physical isolation, and under the constant application and anticipation of human beings, the digitization trend has been transformed into a political rationality. Under the constant application and expectation of human beings, the deeper the digitization trend goes, the more the public and the power rely on it, and the individual human being is forced into an irresistible situation, and the citizen's right to privacy eventually becomes extinct.
On the other hand, PD has made civil and political rights less real. DT is a double-edged sword, while the public enjoys all the conveniences, they have need to bear the negative impacts, all of them have need to face the manipulation of human cognition. Big data technology knows more about them than people themselves. When people access information and express their views through digital media, their personal preferences and political orientations are analyzed and mastered, the more convenient personalized recommendations bring individuals into an information cocoon, and it is difficult for individuals to get away.
Technical manipulation of individual perceptions is achieved through the deliberate and carefully constructed personalization of DT. Such cognitive manipulation, in the political field, is manifested in the manipulation of public opinion, the expression of public opinion manipulated by DT is called the post-truth politics .
A typical example of this is the 2016 election of Donald Trump, in which the British company Cambridge Analytica stole the data of 87 million users from Facebook, and then used the data on the political inclinations of these users, mainly user profiles and information to manipulate the expression of public opinion, profiles, data on likes and comments, and personalized tests to analyze voters’ voting intentions . For voters who have a clear preference for the target candidate, there is no need for informational interventions, but rather the focus is on reinforcing the political preference of the potential voter. The main way to do this is through precise information delivery, i.e. creating an information cocoon for potential voters to promote the target candidate and influence their voting intentions.
The public manipulation can be summarized in three steps: firstly, the classification of political preferences through personal privacy data, secondly, the formation through personalized recommendations, and the last is “guiding voters to cast their votes by carefully selecting and producing content to carry the manipulators’ intentions” . The final election results, which appear to the voters to be the expression of genuine public opinion, are in fact the expression of manipulated civil and political rights.
4.2. PD's Neutrality Becomes an Excuse for Power Abuse
The abuse of power using the superficial characteristics of technological neutrality seems to have found a reliable justification, which is mainly reflected in the fact that technological rationality, or neutrality, hides the power irrationality, the mathematical models and algorithms alienate the discretionary procedures, the lack of regulation of PG, and the inability to define the legal responsibility in three aspects.
Firstly, technological neutrality hides the fact that power and its implementation subject is a limited rationality, the victory of human construction naturally determines the existence of human limitations, “the basic physiological limitations of human beings as well as the cognitive limitations arising from the motivational limitations and their interactions” , which would result in the individual's inability to make decisions in line with the rationality, the decision-making would necessarily include emotions, preferences, and so on, digital technologies, such as Big Data and algorithms perfectly compensate for the physiological deficiencies of human beings, surpassing them in terms of scope of knowledge, breadth of information, and hyper-rationalization, which is easily associated with “controllability, predictability, cost-benefit concepts, and dispassionate logic” , as opposed to emotion, empathy, impulsivity, and surprise .
DT is helping to gradually change the objective conditions that constrain human beings from moving towards full rationality, the traditional assumptions of absolute rationality of human nature do not appear to be idealized, and in the new era, tools are available to move towards an infinite proximity to absolute rationality” . AI provides an optimistic rationale for an individual's ability to deal with complex problems and serve as a reliable analytical tool to give trustworthy and optimal decisions, which individuals are more willing to follow.
Therefore, under the PD’s operation process, technical rationality becomes the best colleague of irrationality , the traditional human-human relationship mode has transformed into human-technology-human relationship mode, the power role has transformed accordingly, the traditional mechanism of “power to people” for the “power to technology to people”, power is hidden behind the curtain of technology.
Automated administration represents the process of making judgements through algorithms that set relevant parameters and goals, and making administrative approvals, executions, and other specific administrative actions. On the surface, this process is the technical rationality to support the automated administrative system to make the relevant standard judgement, the formation of the relevant decision-making (administrative action) is naturally reasonable, the most efficient decision-making, in this set of seemingly perfect process, the power seems to have been a good, legitimate exercise of power, the irrationality of the subject of the power is completely excluded. In fact, DT controlled by PG with a certain bias, this kind of bias is the embodiment of irrationality and covered up by DT’s rationality.
Secondly, DT provides the possibility of differentiation of discretion, which leaves room for power abuse and rent-seeking. Therefore, how to apply discretion in a standardized and reasonable manner is an important consideration for achieving justice in individual cases. Automated equipment based on mathematical logic has become a tool for standardizing and rationalizing discretion and is used in administrative and judicial activities. The risk is that it is precisely its mechanical mathematical logic characteristics that lead to the risk of alienation of discretion.
The realization of AI’s discretion lies in the fact that the parameters or standards generated after data training can correspond to the marked case fact elements. However, there is no clear legal basis for related procedures, such as data selection and purpose setting, this allows AI’s discretion to transcend the limitations of general legal procedures. As the author explained in the previous algorithm’s black box, it makes it impossible for the counterparty to participate in the process of automated discretion. The current law has not formulated relevant systems to protect the rights of the counterparty. Therefore, the realization of legal values, such as fairness and justice, depends on whether the algorithm logic is good or bad, the advantages and disadvantages of the algorithm system operation, and the truth or falsehood of the algorithm learning materials, the algorithm has become the only factor that determines the realization of social fairness and justice. In this process, the discretion of law enforcers transformed into the discretion of algorithms. Compared with the traditional face-to-face administrative enforcement procedures, administrative enforcement after digitization, is hidden, but power still exists, “continuing to exist in an invisible place” .
Finally, supervision and responsibility become a difficult problem. DT’s black box logic is embedded in the power operation, and various institutional mechanisms of traditional power supervision are gradually losing their effectiveness. The specific operation of the algorithm becomes a black box. The algorithm itself does not need and cannot explain its calculation process. For complex algorithms, even programmers find it difficult to understand how it works. With the rapid development of artificial intelligence, the iteration and update of the algorithm itself cannot be fully mastered. In administrative penalties in automated administration, the input of case facts and the results of administrative penalties are visible, but the intermediate processes, including hearings, administrative discretion, and other procedures, are not completed in the visual program. In the application of big data investigation technology, the collection of personal information, the protection of personal privacy, and the scope of personal surveillance in the investigation behavior cannot reflect in the visual program. Due to the non-disclosure, opacity and complexity brought by the algorithm, even professional technicians cannot fully understand the operating logic, how to supervise the operation of power becomes empty at the factual level.
In addition, the legal responsibility after the digitization of power cannot play a legal role. One reason is that the design, operation, and maintenance of automated decision-making includes the participation of administrative law enforcement personnel and technical personnel, as well as the counterparty and other multiple parties. These subjects all play a role. If there is a problem with the automated system and its parameters, who will bear the legal responsibility, or who is the responsible one? Due to DT’s application, “the parties involved in the legal relationship are isolated” , and the traditional legal and illegal responsibility identification and accountability mechanisms and principles cannot be adapted to the human-technology-human model, technological neutrality excludes the counterparty's right to know and serves as a reasonable excuse for shirking responsibility.
5. Two Possible Plans
In recognition of the double-edged sword as a basis, through reasonable institutional adjustment, it is also possible to guide PD as far as possible in the positive direction of promoting social development and safeguarding the rights of citizens. To cope with the PD’s legal risks, the imbalanced configuration of the system of “power-rights” is readjusted by using the digital rights (freedom) of citizens and legal persons to check and balance the technologized power, and a balance between technical rationality and value rationality should be achieved through the establishment of a system of legitimacy tests for the technical operation of power.
5.1. Citizens’ Digital Rights Are Both Public and Private
It is a key task to reconstruct the traditional basic rights of citizens to ensure the free and comprehensive development of citizens in digital social life. China already has a series of laws and regulations such as the “Personal Information Protection Law”, “Data Security Law”, “Cybersecurity Law”, and “Electronic Signature Law”. There is a consensus on the private law attributes of digital rights, and its public law attributes still need to be paid attention to.
Those civil law scholars usually use the relationship between the “Personal Information Protection Law” and the “Civil Code” to “interpret the rights of individuals and legal persons in information processing activities as civil rights” and use this as a basis to establish a private law protection system. The author believes that digital rights not only have the attributes of private rights, but it should be further identified as basic rights of citizens in the sense of public law. Combining “Personal Information Protection Law” with the Constitution, taking information rights as a component of the basic rights of citizens, and combining the private rights protection from the “Civil Code” and the public rights protection from the Constitution with each other, this approach is conducive to getting out of the asymmetric power structure between people and information processors and more conducive to realizing citizens’ dignified digital social life.
The so-called asymmetric power structure refers to the situation in which one party that enjoys an advantage in digital technology has an advantage over the other party in the relationship between international institutions, enterprises and natural people. This unbalanced legal relationship is caused by digital technology, digital monopoly and PD. The complexity of professional knowledge and the economic value and social governance value of DT determine the impossibility of realizing the universal possession, especially in public law legal relations, under the condition of unequal status, the confrontation between individuals and information processors in the asymmetric power structure needs to achieve the balance of powers-rights through the joint force of public law and private law rights mechanisms.
Individuals are digitized in both body and behavior; digital existence is equivalent to material existence. The digital control of individuals is a restriction on personal freedom. Therefore, taking digital rights as basic rights to protect personal freedom is the meaning of “citizens' personal freedom shall not be violated” in Article 37 of the Constitution of the People's Republic of China, the nation has the obligation to protect the personal dignity of citizens in the digital society from being violated.
Public life in digital virtual space promotes the digitization and transparency of people and enables the formation of domination and discipline based on digital technology. This domination mechanism “preemptively recommends and pushes various political, commercial and other information based on citizens’ behavioral trajectories and data inferences, affects our understanding of our own identity by influencing and strengthening our preferences, prejudices and cognitive frameworks, weakens our interactive relationship with the outside world, and inhibits spontaneous behavior and autonomous actions based on identity” .
Digital rights in the sense of private rights enrich and supplement the connotation of public rights, as the implementation of digital rights in public law, they need to be clarified in the relevant provisions of private law, that is, the law clearly stipulates the granting, confirmation and protection of citizens’ specific private digital rights. In this way, the public and private systems will form a joint force to realize the free and comprehensive development of people in digital life.
5.2. To Establish a Legitimacy Review Mechanism for Technology-Embedded Power
It is a basic requirement of the modern legal system that the operation and allocation of power meet the legitimacy test, the operation of power technology must complete the legitimacy test in accordance with the requirements of the Constitution, otherwise, even if the objective effect of digitalization helps to achieve administrative tasks, it will not be recognized in the normative system due to the constitutional condemnation of power legitimacy.
In the macro-institutional framework, clarifying the scenarios and PD’s boundaries is conducive to eliminating the deep intrusion of power technology into natural people. With the vigorous development of digital government construction and the embedding of digital technology into judicial activities, the operation scope of power technology has become wider. This also provides a new opportunity for national and social governance that has brought more significant governance effects. The “health code governance” during the epidemic has demonstrated great governance effectiveness in epidemic prevention. PD has also led to abuse of power due to the generalization without clear legislation on application scenarios and boundaries. A typical example is that Hang Zhou once used “a new gradient health code” that integrates relevant data from electronic medical records, health examinations, and lifestyle management, and can even evaluate health groups such as corridors, communities, and enterprises through big data, this obviously exceeds the reasonable scope of epidemic prevention and control and is an unlimited expansion of citizen data collection. In addition, the use of big data investigation technology has greatly promoted the maintenance of public security and made it possible to prevent crimes in advance. However, this kind of prevention in advance presupposes the idea that “everyone is a potential suspect of crime” .
Therefore, based on the consideration of public interest, big data investigation technology is likely to conduct real-time technical monitoring of the public in addition to crime investigation. These situations show that there is a great possibility that technological power will use legitimate reasons such as public interest to unscrupulously spy on and monitor the public, it obviously deviates from the legal and rational requirements of digital technology in a society ruled by law. It is necessary to clarify the application scenarios and boundaries of digital technology through legislation, enhance the controllability of digital technology, and effectively prevent the infringement of the legitimate rights of the public by the technological power.
After power is technological and digitalized, the formation of decisions is invisible in the hidden layer operation program of digital technology. “Algorithmic decision-making is completed quickly through autonomous analysis within the algorithm, and the entire decision-making process is closed” . Administrative counterparts have no intuitive feeling about this, so they cannot directly participate in the internal decision-making process of the algorithmic decision-making system and have no opportunity to ask the administrative subject to explain the reasons and listen to the counterpart’s statements and defenses. At this stage, the rights set by traditional administrative laws, including the right to participate and the right to supervise, are difficult to achieve, the main reason is the objective technical status quo of the non-display ability of technical operations.
Therefore, PD, to achieve effectiveness between citizens and digital technology owners and users, the focus cannot only be on the openness and transparency of the intermediate operation procedures but also must be set up before digital technology enters the power system. The pre-examination procedure can provide a feasible space for the public to realize their right to know and right to supervise before the system is set up: introduce the public participation process in the system pre-examination procedure, and the subject with information should fully inform the public of the performance, defects, application scenarios, etc. of digital technology, and listen to public opinions; through system pre-examination, a comprehensive evaluation can be carried out before digital technology enters the power system, and a good interaction between technology and power can be achieved through multi-party supervision and evaluation.
6. Conclusion
In the AI’s era, the binding between power and technology is an irreversible trend. PD’s external characteristics and internal logic explain that the structure of power in contemporary society is changing. While humans are enjoying digital life, we are also constantly paying the price and will continue to bear risks. In the face of irreversible trends, the thinking of legal scholars and technical experts seems to be inconsistent. As technical experts, they always pursue technological innovation and iteration, in comparison, jurists are always concerned about gains and losses. When big data, digital life, digital people, and generative AI have been universally recognized, jurists seem to always lag.
However, it must be admitted that it is necessary to keep some calm in the fanaticism wave, especially in the face of national power and its power system. The power system represented by administrative power has not only not paid the price for technological changes but has given power itself the opportunity to hide behind technology and continuously expand its control over citizens and even the whole society. In this process, traditional administrative law theories and administrative procedures seem to be gradually ineffective. Obviously, relying solely on supervision and accountability would not be enough to meet people's various demands. This requires constant thinking about what the law should do and what legal scholars should do in the general demands of society and ordinary people.
Abbreviations

AI

Artificial Intelligence

DT

Digital Technologies

HD

Human Digitalisations

PD

Power Digitalisation

Funding
This paper is supported by the University of Alicante, Spain, as the part of the Programa Redes-I3CE en Investigación y Docencia Universitaria-6298-Universidad de Alicante.
Author Contributions
Yang Yang is the sole author. The author read and approved the final manuscript.
Conflicts of Interest
The author declares no conflicts of interest.
References
[1] Chen Yanting (2022). Legal Control of Assisted Automated Administration. Journal of East China Normal University, vol. 01, pp. 115-213.
[2] Fu Jingyu (2024). Defense dilemma and its response in big data investigation. Journal of Hubei Police College, vol. 37, no. 02, pp. 115-128.
[3] Guo Yuting (2023). Classification of Legal Control of Automated Administrative Decisions Research on the System. Journal of Smart Rule of Law - Collection of Research Papers on Mathematical Law, vol. 01, pp. 157-166.
[4] Guo Zhe (2019). Epistemology of the Investigation Model of Duty Crime in the Big Data Era. Chinese Political and Legal Forum, vol. 03, pp. 42-54.
[5] Hu Minjie (2019). Legal Control of Automated Administration. Administrative Law Research, vol. 02, pp. 56-66.
[6] Kang Lanping (2021). Research on the challenges of generalization of rights and scenario-based responses in artificial intelligence legal research. Journal of Dalian University of Technology (Social Science Edition), vol. 42, no. 01, pp. 98-106.
[7] Liu Hao (2025). The logic and approach of legal protection of credit data in generative artificial intelligence. Science and Technology and Law, vol. 02, pp. 103-113.
[8] Liu Wenqi, Shao Meng (2025). Big data investigation- the dilemma of procurator supervision and its solution. Journal of Sichuan Police College, vol. 37, no, 01, pp. 54-62.
[9] Liu Xiaoqing (2022). From power supervision to rights restriction: the rational dimension of legal regulation of big data investigation. Journal of Chong Qing University (Social Science Edition), vol. 28, no. 02, pp. 220-231.
[10] Liu Zongsheng and Jiang Anna (2024). On the legal attributes of health codes. China Health Law, vol. 32, no. 01, pp. 9-14.
[11] Ma Yanxin (2019). Grading of Automated Administration and Changes in Legal Control. Administrative Law Research, vol. 01, pp. 80-92.
[12] National Bureau of Statistics of China (2024). Digital China Development Report (2023).
[13] Qi Yanping (2018). On the changes of legal scenarios in the era of artificial intelligence. Legal Science (Journal of Northwest University of Political Science and Law), vol. 36, no. 04, pp. 37-46.
[14] Shi Pengpeng (2024). On big data investigation: focusing on the right to information self-determination Line. Research on the Rule of Law, vol. 06, pp. 15-25.
[15] Sun Xiaoxia (2024). How does digital power shape the rule of law — On the logic and mission of digital rule of law. Legal System and Social Development, vol. 30, no. 02, pp. 61-83.
[16] Wang Bin (2023). Legal Control of Algorithms in Automated Administration. Financial Law, vol. 01, pp. 61-75.
[17] Wang Liangjia and Ma Jianbo (2023). Physical law analysis in technology design, application and context - taking health codes as an example. Journal of Jiangsu University of Science and Technology (Social Science Edition), vol. 23, no. 03, pp. 23-30.
[18] Wang Yingwei (2009). An analysis of the existential dilemma of technological rationality. Social Science Front, vol. 02, pp. 243-245.
[19] Wen Hon and Li Fengshan (2023). How does crisis learn drive the continuous innovation of digital governance technology - An investigation based on the evolution of health code. E-Government, vol. 08, pp. 61-72.
[20] Wu Chenchen (2023). Research on Legal Control of Automated Administration. Journal of Harbin University of Commerce, vol. 03, pp. 17-69.
[21] Yan Xiaodong and Enqian Ouyang (2022). Algorithms, power and law: challenges and responses of the times. Chinese Political and Legal Forum, vol. 04, pp. 39-49.
[22] Yang Fan (2025). The legal position of social responsibility in artificial intelligence governance. Administrative Law Research, vol. 03, pp, 129-142.
[23] Yang Ting (2018). On the transformation of the criminal investigation model in the era of big data. Legal and Business Research, vol. 35, no. 02, pp. 25-36.
[24] Yao Jianzong and Gong Zhiwang (2024). Technologicalization of power and its legal risk response in the digital age. Journal of Henan University (Social Science Edition), vol. 64, no. 05, pp. 17-24.
[25] Zhang Bin and Luo Gaoyang (2024). Methods, application and regulation of big data investigation. Journal of Zhou Kou Normal University, vol. 41, no. 06, pp. 91-98.
[26] Zhang Huiyu (2025). Legal Regulation Dilemma and Solution of Computing Power Interconnection in the Era of Artificial Intelligence - From the Perspective of Cooperative Regulation. Journal of Nanjing University of Posts and Telecommunications (Social Sciences Edition), vol. 04, no. 01, pp. 21-49.
[27] Zhang Linghan (2016). The rise, alienation and legal regulation of algorithmic power. Legal Business Research, vol. 36, no. 04, pp. 63-75.
[28] Zhang Qi (2025). Legal risks and responses of personal information protection in the application scenarios of generative artificial intelligence. Journal of Tai Yuan University (Social Science Edition), vol. 26, no. 01, pp. 58-70.
[29] Zhang Xinbao (2015). From privacy to personal information: Theory and institutional arrangements for rebalancing interests. Internet Finance Law Review, vol. 02, pp. 16-21.
[30] Zhang Xinbao (2015). From privacy to personal information: Theory and institutional arrangements for rebalancing interests. China Legal Science, vol. 03, pp. 38-59.
[31] Zheng Shaofei and Yao Jianzong (2022). The legal distribution path of artificial intelligence technology risks under the influence of artificial intelligence technology power. Social Science Research, vol. 03, pp. 110-118.
[32] Zheng Shiming, Song Yue and Zhang Meilian (2023). Crisis learning, attention allocation and policy outbreak: An empirical study based on health code. Chinese Public Administration, vol. 03, pp. 79-87.
[33] Zhou Shangjun (2021). Reconstruction of power mechanisms in digital society. Journal of East China University of Political Science and Law, vol. 24, no. 05, pp. 17-62.
[34] Zou Ziming (2023). Research on the protection of personal information in the use of health codes - Taking the Hang Zhou Red Code Incident as an example. Journal of Hubei University of Science and Technology, vol. 43, no. 04, pp. 41-46.
Cite This Article
  • APA Style

    Yang, Y. (2025). Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization. Journal of Public Policy and Administration, 9(2), 91-99. https://doi.org/10.11648/j.jppa.20250902.14

    Copy | Download

    ACS Style

    Yang, Y. Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization. J. Public Policy Adm. 2025, 9(2), 91-99. doi: 10.11648/j.jppa.20250902.14

    Copy | Download

    AMA Style

    Yang Y. Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization. J Public Policy Adm. 2025;9(2):91-99. doi: 10.11648/j.jppa.20250902.14

    Copy | Download

  • @article{10.11648/j.jppa.20250902.14,
      author = {Yang Yang},
      title = {Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization
    },
      journal = {Journal of Public Policy and Administration},
      volume = {9},
      number = {2},
      pages = {91-99},
      doi = {10.11648/j.jppa.20250902.14},
      url = {https://doi.org/10.11648/j.jppa.20250902.14},
      eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.jppa.20250902.14},
      abstract = {The power’s digitalization in the era of artificial intelligence refers to the deep coupling of technical logic and power logic after digital technology is embedded in the power operation mechanism, thereby promoting a more comprehensive technological transfer in power distribution. With the assistance of technology, power has moved behind the scenes, but the power's ability to control society has been enhanced. The external characteristics of this phenomenon are strong technicality, concealment and intrusion. There are three aspects of the internal logic that led to its emergence: human digital survival has become the norm, the diffusion effect of power is significantly better than its coercive power, and the black box operation space brought by digital technology and people's dependence on digital codes. Using comparative research, empirical research and literature analysis, the challenges encountered by traditional administrative rule of law are explained. The basic principles of modern administrative rule of law established since the 19th century are quietly changing, especially in two aspects: digital technology has eliminated the basic rights of citizens represented by privacy rights and basic political rights, making discretionary power difficult to control and the definition of traditional legal responsibilities difficult. These are challenges to the basic principles of modern administrative law since the 19th century. Controlling or narrowing the legal gaps caused by the digitization of power, combining public and private law protection of citizens' right to information, and establishing a legality review system for technology operations centred on the controllability of digital technology and public participation should become a feasible approach.
    },
     year = {2025}
    }
    

    Copy | Download

  • TY  - JOUR
    T1  - Legal Risks of Digital Governance in the AI’s Era --- Concerns About the Power’s Digitization
    
    AU  - Yang Yang
    Y1  - 2025/04/29
    PY  - 2025
    N1  - https://doi.org/10.11648/j.jppa.20250902.14
    DO  - 10.11648/j.jppa.20250902.14
    T2  - Journal of Public Policy and Administration
    JF  - Journal of Public Policy and Administration
    JO  - Journal of Public Policy and Administration
    SP  - 91
    EP  - 99
    PB  - Science Publishing Group
    SN  - 2640-2696
    UR  - https://doi.org/10.11648/j.jppa.20250902.14
    AB  - The power’s digitalization in the era of artificial intelligence refers to the deep coupling of technical logic and power logic after digital technology is embedded in the power operation mechanism, thereby promoting a more comprehensive technological transfer in power distribution. With the assistance of technology, power has moved behind the scenes, but the power's ability to control society has been enhanced. The external characteristics of this phenomenon are strong technicality, concealment and intrusion. There are three aspects of the internal logic that led to its emergence: human digital survival has become the norm, the diffusion effect of power is significantly better than its coercive power, and the black box operation space brought by digital technology and people's dependence on digital codes. Using comparative research, empirical research and literature analysis, the challenges encountered by traditional administrative rule of law are explained. The basic principles of modern administrative rule of law established since the 19th century are quietly changing, especially in two aspects: digital technology has eliminated the basic rights of citizens represented by privacy rights and basic political rights, making discretionary power difficult to control and the definition of traditional legal responsibilities difficult. These are challenges to the basic principles of modern administrative law since the 19th century. Controlling or narrowing the legal gaps caused by the digitization of power, combining public and private law protection of citizens' right to information, and establishing a legality review system for technology operations centred on the controllability of digital technology and public participation should become a feasible approach.
    
    VL  - 9
    IS  - 2
    ER  - 

    Copy | Download

Author Information
  • Law School, Shen Yang Normal University, Shenyang, China; Legal Culture Centre, Shen Yang Normal University, Shenyang, China; Global Innovation Law and Policy, University of Alicante, Alicante, Spain; Legal History Centre, University of Alicante, Alicante, Spain

  • Abstract
  • Keywords
  • Document Sections

    1. 1. Introduction
    2. 2. PD’s External Characteristics
    3. 3. PD’s Inherent Logic
    4. 4. PD’s Legal Risks and Analysis
    5. 5. Two Possible Plans
    6. 6. Conclusion
    Show Full Outline
  • Abbreviations
  • Funding
  • Author Contributions
  • Conflicts of Interest
  • References
  • Cite This Article
  • Author Information