[author]Huang Pei

[content]

Administrative regulatory pathways in the metaverse: a framework analysis

Huang Pei

Associate Professor, School of Law, Tongji University

This article was published in China Legal Science, No. 6, 2022.

Abstract: Metaverse is a new generation of full-fledged Internet form, and users, data, and algorithms are the three important pivots of social and economic activities in the era of meta-universe, and there are many risks in all three, including: the risk of hindering in-depth cooperation among people due to the anonymisation of meta-universe users' identities, the risk of infringing on users' personal information and privacy, the innovation power of market entities, and the security of national data due to the agglomeration of meta-universe data, the risk of re-centring of information and data, monopolisation of economic activities, and manipulation of social concepts due to the power of meta-universe algorithms. These risks are difficult to be resolved by market forces on their own, and need to be prevented by the power of administrative regulation. Specific administrative regulation paths include the construction of a unified meta-universe identity authentication platform, the implementation of identity authenticity regulation such as access and single identity authentication system for organisational users and multi-identity authentication system for individual users, the implementation of mandatory data desensitisation of user data, the restriction of meta-universe platform enterprises to engage in inter-industry competition with business data, the standardisation of cross-border data flow and other data security regulations, as well as algorithmic transparency, algorithmic ethics and algorithmic accountability regulations.

Keywords: metaverse; administrative regulation; identity anonymisation; data clustering; algorithmic empowerment

Catalogue:

I. Question

II. Identity anonymisation risk and authenticity regulation in the metaverse

III. Data clustering risk and security regulation in the metaverse

IV. Algorithmic empowerment risk and the regulation of goodwill in the metaverse

V. Conclusion

1. Question

Metaverse, in a broad sense, refers to immersive three-dimensional virtual worlds in which people can interact with each other through their avatars, free from the physical constraints of the real world. In short, it is "a digital virtual society with reality". The concept of the metaverse was born in 1992, and the year 2021 saw an explosion of metaverse construction, also known as the Year of the Metaverse. At the beginning of this year, sandbox game platform company Roblox became the first company to include the concept of metaverse in its prospectus and successfully listed on the New York Stock Exchange, with a market capitalisation of more than $40 billion on the first day, which was highly sought after by the capital market. Chip giant Nvidia has announced a simulation and collaboration platform called Omniverse Enterprise, dedicated to helping other companies and individuals replicate real-world digital twins in cyber-virtual spaces. Internet giant Facebook has taken the plunge and renamed its group of companies "Meta", fully embracing the concept of metaverse in order to enhance the company's external image and convey its determination to transform its business model. China's ByteDance spent more than 5 billion yuan to acquire virtual reality (VR) startup Pico, and began to enter the field of meta-universe. Tencent has also applied for the registration of nearly 100 metaverse-related trademarks and invested in many metaverse-related enterprises. Other famous Internet companies such as Alibaba, Microsoft, Amazon, Baidu, etc. have also begun to make efforts in this field, making the metaverse the hottest topic.

Meta-universe is considered to be the third generation of the Internet form, which highly integrates the existing big data technology, cloud computing technology, VR/AR and other interactive technologies, artificial intelligence algorithms, blockchain technology, Internet of Things technology and a series of high-tech, connecting these high-tech "into a line", to form a virtual world intertwined  with the real world. As the meta-universe represents the innovative development direction and high-level form of science and technology, and is the concentrated manifestation of a country's scientific and technological innovation ability, which has huge economic value and is of great significance to improve national competitiveness, the metaverse not only has strong attraction to all kinds of market players, but also should be the strategic goal of future development set by the national level, and become the forward direction of innovative scientific and technological development.At present, the metaverse is in its infancy, and just as the experience of the early development of the mobile Internet has shown, the early development of the metaverse needs to be driven by the strong impetus provided by various market players competing with each other on the basis of economic interests, and market forces will dominate. However, as risks always follow behind the high density and complexity of the digital world, we cannot ignore the risks that may accompany the development of the metaverse by market forces. These risks may not be obvious at the beginning of the metaverse, but they will always be there, and will continue to emerge as the metaverse grows, and even eventually become negative forces that hinder the healthy development of the metaverse. Therefore, there is a need for a force different from the market to prevent these risks before they occur, and administrative regulation thus becomes a necessary choice.

Administrative regulation of the metaverse is in fact the regulation of risks in the process of development of the metaverse, that is, through the establishment of a professional administrative regulatory body, to assess and monitor the risks of the metaverse that may cause public harm and to eliminate or mitigate the corresponding risks by means of formulating rules and supervising the implementation of the rules. It belongs to the problem of choosing the way of metaverse governance and is one of the most important legal issues in metaverse governance. The government should properly delineate the necessary path of administrative regulation, so that while maintaining market forces in building the meta-universe, it can, through continuous supervision and management of market players, prompt them to identify their own problems, formulate countermeasures, and strengthen their own internal governance, so as to adequately reduce the risks associated with the development of the meta-universe. Then, in such a nascent period of meta-universe, in what aspects should the administrative regulatory authorities implement the necessary risk regulation of meta-universe? In this regard, this paper will provide a framework analysis of this core issue based on the following diagram:

Figure 1 Framework diagram of the administrative regulation of the metaverse

As shown in the figure above, this paper selects three aspects: user (identity), data, and algorithms to explore the administrative regulation path of the metaverse. These three are the three important fulcrums for building social and economic activities in the metaverse era: users are the main players in the metaverse. In the final analysis, the metaverse is still to serve the users who are active in it, promote in-depth cooperation between users through the "power of the network", and promote the in-depth development of the social economy. Data is the basic factor of production in the metaverse era. As a high-level form of digital economic development, the metaverse realizes the interconnection of all things after digitization, and generates massive data at all times, which is the basic production factor of the social and economic development of the metaverse. Algorithms are the basic governance tools for the operation of the metaverse. Algorithms of pure technical significance have developed into algorithms of legal significance in the era of the metaverse, and the function of algorithms has also developed from a simple technical product to a basic governance tool of the metaverse, which has irreplaceable and important functions. However, all three have potential risks that hinder the socio-economic development of the metaverse era: the anonymization of the identity of metaverse users will hinder the in-depth expansion of cooperation between users; The agglomeration of data in the metaverse will pose a serious threat to personal information and privacy, the innovation momentum of market players, and national security. The power of algorithms in the metaverse will lead to the use of the recentralization of information and data by metaverse platform companies to achieve the monopoly of economic activities and the manipulation of social concepts. These potential risks are difficult to be effectively constrained by market forces, and even market forces themselves are the reason for the existence of these risks, so it is necessary to introduce the power of administrative regulation to prevent and resolve them.

Based on the above three fulcrums, this paper will discuss the risk of identity anonymization and the corresponding regulation of identity authenticity in the metaverse, the risk of data agglomeration in the metaverse and the corresponding regulation of data security, and the risk of algorithmic power in the metaverse and the corresponding regulation of algorithmic goodness, in order to provide some intellectual contribution to the planning of the path of administrative regulation of the metaverse.

2. Identity anonymization risks and authenticity regulation of the metaverse

The most important function of the metaverse for human society is that it can provide a new way for people to achieve cross-regional and cross-border cooperation, and maximize the cooperative relationship between people. In the virtual space of the metaverse, users can get rid of the physical constraints of the real world and interact with other users through virtual avatars to achieve win-win cooperation. However, the anonymization of users' identities in the metaverse will hinder the generation of trust, which in turn hinders the formation of in-depth cooperative relationships between users, and may even release the "malignancy" in users' hearts and carry out anti-productive activities under the cover of identity anonymization. This will not only disrupt the order of the metaverse itself, but may also affect the social order of the real world through the close connection between the metaverse and the real world. Therefore, it is necessary to implement identity authenticity regulation for Metaverse users through administrative regulatory authorities to ensure that the legal system can actually restrain the behavior of Metaverse users, and provide a foundation of trust for users to reach cooperation through Metaverse platforms.

2.1The risk of identity anonymization in the metaverse

The establishment of cooperative relations between people is extremely important for the survival and long-term development of society, and even the existence of society itself is the existence of cooperative relations between people. On Robinsonian islands, there is no cooperative relationship between people, and there is no social existence. In the real world, modern industrialized societies mainly rely on the legal system established by sovereign states to facilitate the achievement of cooperative relations between people. Because in industrialized society, with the fine division of labor in large-scale production and the convenience of transportation, the mobility of personnel has increased significantly, and the degree of strangeness between people has deepened, forming many regional stranger societies or semi-acquaintance societies. Under the asymmetry of information, people cannot reasonably predict each other's behavior, and it is difficult to generate the necessary trust, and it is difficult to achieve cooperative relations. As a result, the importance of the legal system established by sovereign states is highlighted. The legal system is a collection of behavioral rules formulated by sovereign states, and with the support of the state's coercive power, it can build an ideal legal order framework and provide people with authoritative rules of behavior that must be followed together. When people follow these common rules of behavior and trust that others will follow them, the barriers to unpredictability are broken, trust between strangers is created, and partnerships are formed.

However, there is an important premise for the realization of this function of the legal system, that is, the subject of public power responsible for the implementation of the law must accurately grasp the true identity of the people in society. Because only by knowing people's true identities can we ensure that the coercive power of the state reaches every individual in society - if an individual violates the rules of conduct set by the law, he will be effectively sanctioned and bear the corresponding legal responsibility. If people's true identities are not grasped, then once the rules of conduct set by the legal system are broken, it will be impossible to accurately locate the specific subject of legal responsibility, and the coercive power of the state will not be truly reflected. At the same time, only by "believing" in the authenticity of the identity of the partner will people believe that when the partner violates the rules of conduct, the coercive power of the state can directly force the partner to abide by the corresponding rules of conduct and bear the corresponding legal responsibility, so that their own rights and interests can be effectively protected. If the identity of the partner is not true (i.e., anonymized), then it is difficult for people to believe that the other party's future behavior can be effectively constrained by legal coercion, and they may think that their own interests are not effectively protected, and it will be difficult to achieve cooperation. In other words, the authenticity of identity is one of the prerequisites for the legal system to actually bind people's behavior, and people's mutual belief that each other's behavior can be effectively constrained by the legal system is a necessary condition for trust, and trust is the basis for reaching a cooperative relationship. Therefore, the authenticity of identity becomes a necessary prerequisite for cooperation between people, which enables people to reasonably anticipate the possibility of future behavior of the partner, establish mutual trust, and reach a cooperative relationship between them.

No matter what form the metaverse takes, it is essentially a society. The metaverse cannot exist in isolation from the real world, it should be an extension of the real world rather than a parallel domain, an expansion rather than a replacement. In terms of its expansion function to the real world, the main significance of the existence of the metaverse lies in its ability to transcend the limitations of physical space in the real world to the greatest extent, and establish cooperative relationships between people through virtual embodiments in a fully realistic virtual environment, thereby promoting social development in the real world and enhancing the total welfare of society in the real world. If it cannot effectively promote the cooperative relationship between people and enhance the well-being of people in the real world, then the meaning of the existence of the metaverse may only be to simply provide the function of games and become a large-scale game console, rather than becoming the support and direction of the long-term development of the future society.

The main body of activity in the metaverse is the virtual avatar of the user in the real world, which can be called a "digital human", which mainly includes two types: virtual digital human and holographic digital human. Virtual digital humans are pure avatars created by digital technology in the metaverse, and are usually actively shaped by organizational users (mainly various companies and enterprises) for business purposes. The holographic digital human is the corresponding digital 3D image of individual users in the metaverse, and users can capture their subtle movements through wearable extended reality devices, digital twins technology, and "interactive reality" (Cross-Reality) technology and synchronize their subtle movements in the metaverse in real time to achieve the realistic effect of mirroring. At the same time, the scene information of the holographic digital human in the metaverse can also be synchronously conveyed to the user, and it is not limited to the visual information and auditory information transmitted by the traditional Internet, but also includes tactile information, olfactory information and even taste information, so that users can feel an immersive experience that is almost indistinguishable from the real world. This immersive experience, combined with the variability of digital technology, allows users to have multiple different identities in the metaverse, deeply experience different roles, experience real life, work, and entertainment scenes, and even build their own small world in the metaverse.

Whether it is a virtual digital human or a holographic digital human, it is closely related and inseparable from metaverse users. Virtual digital humans are virtual products designed and created by metaverse users of the organization, and the behavior pattern is completely controlled by the user's pre-set design at the time of design, and its behavior essentially reflects the user's thoughts and thoughts. The holographic digital human is a true image of the individual metaverse user, and every move is controlled by the user, and there is no independent holographic digital human separated from the metaverse user. Therefore, the regulatory function of the legal system in the real world on user behavior can naturally radiate to the behavior of digital humans in the metaverse.

However, in the metaverse, digital humans under the control of digital technology can have multiple identities, and behind this multiple identities is the anonymization of users in the metaverse, and the digital 3D model is like a real and three-dimensional mask that hides the user's true identity. As shown in the above analysis, anonymization will weaken the coercive power of the legal system on users in the metaverse, and at the same time, it will also lack sufficient trust between these users who are connected through the metaverse - because it is difficult for users to trust an anonymous partner who is not subject to the coercion of the legal system of a sovereign state - As a result, users can only engage in some individual or simple cooperative activities by manipulating digital humans in the metaverse (such as playing alone or playing in a team with other users in the metaverse), and it is difficult to form cooperative relationships with significant interests (such as signing real estate sales contracts in the metaverse). As a result, the most important function of the metaverse, which is to facilitate cooperation between people, cannot be realized, and the metaverse may become a pure game console.

1.2 Regulation of the authenticity of the identity of the administrative regulation organs

It is precisely because of the above risks associated with the anonymization of identities in the metaverse that there must be a force that can ensure the authenticity of the identities of metaverse users, at least make metaverse users believe in the authenticity of the identities of other users, establish a certain mapping relationship between the user's avatar and the real identity, and effectively realize the investigation and implementation of legal responsibility. At present, platform companies building the metaverse cannot effectively provide this power, because metaverse platform enterprises naturally have a tendency to maximize their own interests, and the anonymization of user identities will not bring obstacles to the metaverse platform enterprises to obtain more economic benefits (for example, the anonymization of users in metaverse games does not prevent the fees of metaverse platform enterprises), but may expand the economic benefits of metaverse platform enterprises (for example, metaverse games that do not require real-name registration can always attract more users). Therefore, it does not aim to foster partnerships between metaverse users. In other words, the anonymization of the identity of metaverse users will essentially lead to a decrease in the willingness of users to cooperate in depth and hinder the long-term development of the social economy, but it will not harm the immediate interests of metaverse platform enterprises, and it will have no incentive to solve this problem.

The long-term development of the social economy itself is actually a public good, with the characteristics of non-competition and non-exclusivity, which is also the main reason why it cannot be fully entrusted to the main body of the market. As a result, the power of administrative regulation reflects its own comparative advantage. For a government, the long-term socio-economic development of the area under its jurisdiction essentially determines the legitimacy of its existence and is also a solid foundation for consolidating its dominant position, so the government has a strong driving force to promote the long-term socio-economic development of the area under its jurisdiction. Promoting in-depth cooperation between people in a wider range of fields through administrative regulation and continuously promoting upward social and economic development has become an option compatible with government incentives. In order to prevent the anonymization of the identity of Metaverse users from hindering the realization of in-depth cooperative relations between people, administrative regulatory authorities can enhance mutual trust between users through regulatory tools that locate the true identities of Metaverse users, so that users can continuously enhance in-depth cooperation through the Metaverse on the basis of this trust. Specific administrative regulatory paths include:

First, build a unified national metaverse identity authentication platform. The national unified identity authentication platform is an effective way to carry out trusted identity authentication on the Internet. The unified identity authentication platform is built on the credibility of the government, which can provide authoritative identity authentication information, and can use the existing identity authentication database to improve construction efficiency and reduce construction costs. Judging from the current construction of the metaverse, there are many metaverses built by different platform companies. For the sake of their own interests, the metaverses built by different platform companies often implement user identity isolation between each other, and the user identities in one metaverse usually cannot be seamlessly connected to another metaverse, requiring users to re-register, authenticate, and log in. This requires a unified identity authentication platform that can unify all metaverses and provide authoritative identity authentication for users in different metaverses. The metaverse identity authentication platform built by the government is the most suitable choice, because the government has the willingness, authority and ability to coordinate various platform enterprises, can efficiently provide centralized identity authenticity authentication, and realize the identity exchange of users between various metaverses. There may be a view that the blockchain technology in the metaverse can verify the authenticity of identity information, so there is no need for the government to build a unified metaverse identity authentication platform. This view ignores the fact that although the current blockchain technology in the metaverse can provide decentralized identity information verification, this verification is only a "completeness" verification of identity information, that is, it can only ensure that the identity information submitted by the user is not tampered with after submission, and cannot do anything to determine whether the identity information itself is "true" after the user initially submitted. Therefore, blockchain technology cannot replace centralized identity authenticity, and the two are rather complementary to each other.

Second, the metaverse access system and single identity authentication system will be implemented for organizational users. The so-called organizational users refer to impersonal entities such as companies and social groups. There are a large number of such users in the metaverse, who use the platform functions of the metaverse to engage in various social and economic activities and expand the scope of their business activities. Since the main purpose of organizational users entering the metaverse is to engage in social and economic activities, and they usually have stronger activity ability than users as individuals, their behaviors can have a more far-reaching impact in the metaverse than the behaviors of natural persons. Once an organizational user breaks the rules of behavior in the metaverse, the damage caused will be even greater. Therefore, the administrative regulatory authority should implement an access system for organizational users in the metaverse, and organizational users need to obtain administrative permission from the administrative regulatory authority before they can enter the metaverse to carry out social and economic activities. Organizational users should implement a single identity authentication system, that is, organizational users can only use the real organization name in the real world in the metaverse, and can only correspond to a fixed and public avatar in the metaverse, so as to prevent them from engaging in anti-productive activities through anonymization and disrupting the order in the metaverse.

3. Data agglomeration risks and security regulations in the metaverse

In modern society, the importance of data has become increasingly prominent, and it has become an indispensable factor of production in social and economic activities, and it is also the basic material for building the metaverse. However, the rise of the metaverse is accompanied by an increasingly dense and agglomerated amount of data. This massive amount of data, combined with the ever-evolving big data analysis technology, enables metaverse platform companies to have full control over the data of user activities. For individual users, this means that metaverse platform companies can have an unobstructed view of their personal information, and can accurately grasp their behavior dynamics, and even accurately predict their thinking trends. For organizational users (especially commercial companies), this means that the business information and effective business models obtained by paying the cost of trial and error may be easily copied by metaverse platform companies, forming unfair competition and hindering market innovation. At the same time, if the massive user data gathered on the metaverse platform is consciously developed and utilized, it will also pose a major threat to national security. These risks cannot be resolved by the power of the market itself, and the data security regulation of the metaverse must be implemented through the power of administrative regulation, so as to protect users' personal information and privacy, maintain the innovation momentum of market players, and ensure the data security of the country.

(1) The risk of data agglomeration in the metaverse

As a new generation of the Internet, the metaverse will become one of the most important places for people's daily activities. Since the metaverse itself is built through digital technology, the user's activity information in the metaverse can be easily recorded by the system in the form of data. In addition, wearable extended reality devices can upload almost all of the user's biological characteristics to the metaverse, and record all the user's behavior information in the metaverse, including activity tracks, speech, movements, body temperature, heart rhythm, pulse and even neural activities. Some XR devices even require data to be scanned and recorded about the user's surroundings in the real world. As a result, the volume of user data formed and stored in the metaverse far exceeds the volume of data collected through other digital technologies, becoming a true big data. The metaverse has become a source of continuous production of user-related information, collecting and storing almost all of the user's information and data, making the user "Quantified Self" or "Measurable Self". On the one hand, these user big data can provide data production factors for the continuous improvement of the metaverse, so that the authenticity of the metaverse is closer and closer to the real world (for example, it provides data raw materials for the artificial intelligence self-learning of virtual digital humans, so that virtual digital humans are closer and closer to real biological humans, and have the ability to learn and evolve actively); But on the other hand, these user big data will also be actually controlled by the metaverse platform enterprises, resulting in the following risks:

First, it infringes on the user's personal information and privacy. Metaverse platform companies are market players that pursue profit maximization, and their main purpose in building and operating the metaverse is to make profits. In order to maximize their own interests, these metaverse platform companies will inevitably maximize the use of user personal information data collected and stored in the metaverse. Due to the serious information asymmetry between metaverse platform enterprises and users, the former can use these information big data in a way that is unknowable to metaverse users, and even trade and transfer these information big data in order to seek more revenue, and the possibility of users knowing about this is not high. In fact, even if users are aware of the infringement of their personal information and data by metaverse platform companies, under the disparity in strength, users, as individuals, lack the ability to hold the metaverse platform companies accountable and cannot effectively protect their own rights and interests.

What is more noteworthy is that the use of big data technology by metaverse platform companies can accurately predict the possibility of users' future behaviors, and even infer what users think, so as to deeply intervene in the user's personal privacy space. The main function of big data is prediction, that is, to predict the probability of things happening through the mathematical algorithm operation of massive information data - although there are many types of big data, but theoretically speaking, if there is enough big data as a sample, combined with data analysis technology, it can make more accurate predictions about the future. For example, the likelihood of predicting future climate change based on data on past climate conditions in a particular area, or predicting the likelihood of which means of transportation a natural person will choose in the future based on data on his or her past movements. As mentioned above, the volume of user information data collected by Metaverse platform companies in the Metaverse far exceeds the volume of data collected by ordinary data technology, so with the mining of highly developed big data analysis technology, Metaverse platform enterprises can more accurately predict users' future behavior choices. What's more, through the in-depth analysis of user information big data, metaverse platform companies can also deduce users' thinking activities and grasp users' thinking dynamics. Because the user's thinking will always be manifested through its behavior, there is no simple behavior that is divorced from the activity of the mind. In the real world, user behavior often disappears after implementation, making it difficult to repeat observation. Even if it is recorded by electronic monitoring equipment, the volume of data is always limited, from which only the user's simple thinking dynamics can be analyzed. However, in the metaverse, every user's every move, even every breath and heartbeat, in the metaverse will be recorded exactly and converted into data that can be analyzed. Through the analysis of these behavioral information data, in theory, metaverse platform companies can accurately infer the main thinking activities of users and deeply intervene in the user's individual privacy space.

Second, it hinders the innovation momentum of market players. The metaverse is a high-level Internet platform, in which in addition to individual users, there will be many organizational users, especially corporate users as market entities. These corporate users are the main force in the development of economic activities in the metaverse. In the foreseeable future, perhaps all companies will have their own place of operation in the metaverse (just as almost all companies now have their own web page) or even use the metaverse as their main place of operation. Unlike the traditional Internet, the metaverse provides a virtual environment that allows companies to conduct almost all of their business communication activities, not much different from those conducted in the real world. Even based on highly developed interactive reality technology, even manufacturing companies can directly remotely and accurately control machines and equipment in the real world through the virtual environment provided by the metaverse, and complete all manufacturing processes in the metaverse.

Third, it threatens the country's data security. The highly agglomerated data collected by metaverse platform companies based on the metaverse may also pose a potential threat to national data security. Because of the transnational nature of the metaverse, many of the platform companies that build the metaverse are large international companies. Once these multinational metaverse platform companies collect a sufficient amount of information and data through the metaverse, they can predict a lot of information related to national security. A typical example is that Didi has collected a large amount of travel data of staff of central and state agencies through travel app software, and combined with these travel data to analyze the behavior patterns of staff of specific national ministries and commissions, including a lot of information that may be related to national security. At present, commercial companies such as Didi only have big data in specific segments, which may already have potential threats to national security, not to mention that the volume of user data collected in the metaverse is all-round, far exceeding the amount of information and data obtained by commercial companies in such segments, and the potential threats to national security that may exist are even more far-reaching. This is highlighted by the global Internet monitoring project revealed in the "Prismgate incident".

3.2 Data security regulations by administrative regulation organs

It is precisely because of the above risks in data agglomeration in the Metaverse, and these risks are inherent in the essential characteristics of Metaverse platform enterprises as for-profit organizations, which are difficult to prevent and dissolve through market forces, so it is necessary to implement data security regulation on the Metaverse through the power of administrative regulation to ensure that the agglomeration data formed and stored in the Metaverse can play a positive role and avoid causing actual damage. The specific administrative regulation path can start from the following three aspects:

First, implement mandatory user data masking. There is a necessary premise for the high concentration of user information data in the metaverse to produce actual infringement of the rights and interests of individual users, that is, the user's information data must be able to be associated with the real identity of the user. Because only the information data can be confirmed as the data of a specific identity user will lead to the infringement of the rights and interests of the specific user, otherwise the information data of individual users is just the information itself. For example, in the case of "virtual harassment" that has occurred in the metaverse, if the real identity information of the victim in the real world is not known, the impact of such an incident on the victim is limited to a specific scope, and the damage is relatively limited. Once the victim's real identity information in the real world is disclosed, and the information about the virtual harassment incident is publicly linked to the real identity of the specific victim, then the victim may have more shame and mental pressure than when the real identity information is concealed, and thus suffer more harm.

As for what kind of information data can directly identify the real identity of individual users, the criterion of "a motivated intruder" can be used to determine what kind of information data can be used. "Motivated intruder" refers to "having the reasonable ability of an ordinary person to establish a reliable connection between personal data and a specific individual through technical means (such as the Internet, libraries, various public archives, etc.) that can be used by ordinary people with a clear purpose, so as to directly identify a specific individual". It cannot be a data analysis expert with a high degree of professional technology, nor can it be a subject with high-tech data analysis tools, and the data analysis ability should be between the general public and technical experts. If a motivated intruder is able to identify the true identity of the relevant individual through specific data, then the data is information data that can directly identify the real identity of the individual user. Under the constraints of this standard, the highly concentrated individual user information data in the metaverse can be used as a production factor for the construction of the metaverse while minimizing the risk of infringement of users' personal information and privacy.

Second, restrict the competition of metaverse platform companies. As the main platform for future social and economic life, the metaverse will become the main place for all kinds of companies and enterprises to engage in economic activities, and the massive business data generated by it contains huge value, and if properly used, it will become an important source of technological innovation and business innovation. However, the metaverse platform companies that master these business data are themselves for-profit economic organizations, although they may also carry out technological innovation and business innovation out of the need to realize their own interests, but innovation is a high-risk investment activity that requires huge costs to be paid in advance, and there is a strong uncertainty. Therefore, when Metaverse platform companies find that they can quickly replicate the mature business models of companies on the Metaverse platform through the analysis of these massive commercial data, avoid the trial and error costs caused by innovation, and obtain certain benefits, it is difficult to hope that Metaverse platform companies can exercise self-restraint and not use these commercial data to carry out unfair competition with companies on the Metaverse platform. As mentioned above, this unfair competition between metaverse platform companies and metaverse platform companies will eventually lead to the continuous weakening or even death of the innovation momentum of companies on the platform. Therefore, there is a need for administrative regulation as a non-market force to prevent unfair competition between metaverse platform enterprises and metaverse platform companies. The administrative regulatory authorities should strictly restrict Metaverse platform enterprises from carrying out the same business as those operated by companies on the platform, and require Metaverse platform enterprises not to use the commercial data they obtain to compete with companies on the platform. The ideal goal to be achieved by the administrative regulation may be to shape the metaverse platform enterprises into a purely service-oriented Internet platform, so that it can provide a real virtual world of in-depth cooperation and fair competition for various economic activities, and provide various conveniences and possibilities for scientific and technological innovation and business innovation.

Third, regulate the cross-border flow of user data. As a new generation of the Internet, the metaverse naturally has the characteristics of cross-border, and the metaverse built by many multinational companies is aimed at different people around the world, so the problem that arises is that the data formed by the activities of users in the metaverse in their own countries may flow across borders, and there are huge national security risks in this flow. For example, the European General Data Protection Regulation (GDPR) sets a strict regulatory approach to cross-border user data flows, and provides for the processing of data by companies that provide goods or services to EU residents through a unique "long-arm jurisdiction" clause. China has also established a basic legal framework for cross-border user data flow (for example, Article 37 of the Cybersecurity Law stipulates that personal information and important data collected and generated by critical information infrastructure operators in the course of their operations within the territory of China shall be stored within the territory, and if it is truly necessary to provide it overseas for business needs, a security assessment shall be conducted; Article 34 of the "Map Management Regulations" stipulates that Internet map service units shall set up servers that store map data within the territory of China); However, in general, these laws and regulations have very broad provisions on the cross-border flow of user data, and do not stipulate the scope of important data, security assessment standards, and flow procedures for the cross-border flow of user data, nor do they establish a detailed way to classify and regulate the cross-border flow of user data. In addition, the data regulatory departments of different industries are fighting their own battles, only supervising the cross-border flow of data in their own industries, and lacking an effective cooperation and coordination mechanism. These regulatory issues related to the cross-border flow of user data will cause many risks related to national security and social security even in the traditional Internet era, and such risks will continue to intensify in the era of the metaverse, and the need to improve the regulatory system for the cross-border flow of user data is more prominent.

4.he risk of algorithmic power and the regulation of goodness in the metaverse

Everything in the metaverse exists digitally, which makes algorithms play a fundamental role in the operation of the metaverse. Through the control of algorithms, metaverse platform companies can guide and control the activities of metaverse users at the most fundamental level, forcing users in the metaverse to obey the will of metaverse platform enterprises, thus generating "algorithmic power". There is a risk of alienation of algorithmic power, which will lead to the possibility of infringement of the rights and interests of metaverse users. Based on the high degree of information asymmetry, the risk of infringement of rights and interests is difficult to prevent by the users of the metaverse themselves, but can only be resolved by the power of administrative regulatory organs to guide the algorithm for good.

4.1The risk of algorithmic empowerment of the metaverse

The meaning of algorithm in computer science technology refers to "a limited, deterministic, effective and suitable for computer programs to achieve a problem-solving method", which is essentially a mathematical process of solving problems, and is a specific thinking path and method. However, in the context of big data in the metaverse era, this purely technical logic process, combined with massive information big data, can produce the dominant power of what Max Weber called "the possibility of imposing the will of the individual on the actions of others", so that the algorithm has the power attribute in the Weber sense. This is because people always decide their behavioral choices through the information and data they obtain, and the activities of metaverse users are also a process of continuously obtaining information and data through the metaverse and making behavioral decisions. The cross-regional and even cross-border characteristics of the metaverse have expanded the breadth and depth of information and data that can be obtained. However, the ability of metaverse users to process information data is limited after all, and their physiological limitations (such as energy allocation, health status, life expectancy, etc.) make it difficult for massive information data to be efficiently analyzed in a short period of time, so algorithms have become a powerful tool to assist metaverse users in analyzing information data and generating efficient decisions. By simulating the thinking process of biological humans in the real world, the algorithms in the metaverse can continuously and efficiently analyze massive information data without restrictions within the technical boundary, and combined with artificial intelligence technology, they have a certain ability of independent learning and evolution, which can assist or even replace metaverse users to make behavioral decisions, thereby greatly improving the decision-making efficiency of metaverse users.

However, the algorithm is a double-edged sword, while assisting Metaverse users to improve the efficiency of behavioral decision-making, it can permutate, combine, filter and push the massive information data in the Metaverse according to specific programs through preset mathematical logic programs. This preset mathematical logic program implies the value judgment of the metaverse platform enterprises themselves, and essentially integrates the preferences of the metaverse platform enterprises. Metaverse platform enterprises can guide the behavioral decision-making and choice of Metaverse users by adjusting the output of information and data results, and secretly manipulate the cognition and preferences of Metaverse users by weaving different forms of "information cocoon" in the sense of Keith Sunstein, so that Metaverse users can consciously obey the will of Metaverse platform enterprises, thereby generating dominating power over the behavior and decision-making of Metaverse users. This makes algorithms a de facto form of power, which some scholars call information dominance.

The domination of algorithmic power is a kind of invisible domination, which is different from the traditional explicit domination of public power, which does not show its own will through directly observable power operation signs, but by manipulating the information and data content that Metaverse users rely on to make behavioral decisions, and "moistens things silently" by virtue of its own architectural advantages to guide the action direction of Metaverse users. As a result, the obedience of Metaverse users to the will of Metaverse platform enterprises is a kind of unconscious obedience, and even Metaverse users themselves will not perceive that this is an obedience to the power of algorithms, but will think that their own behavior, decision-making and choices are out of autonomous consciousness. In other words, the domination of algorithmic power does not come from the unilateral suppression and coercion of coercive forces, but from the unilateral invisible control of information data and processing programs by metaverse platform companies. The obedience of metaverse users is not obedience out of fear, but self-identification in a high degree of information asymmetry. Therefore, the operation of algorithmic power is more subtle and imperceptible than that of traditional state public power, and its rigidity is weaker but more resilient. This hidden information dominance of algorithms may lead to three risks in the metaverse:

First, the recentralization of information and data. The wide application of blockchain technology is an important technical feature of the metaverse that is different from the traditional Internet, and the digital environment built based on blockchain technology in the metaverse is usually considered to have important characteristics of "decentralization". Blockchain technology enables the information data in the metaverse to be stored in a distributed manner, and each storage node cannot modify the information data independently without being known by other nodes, so it can fully ensure that the information data is not tampered with. Therefore, Decentralized Autonomous Organizations (DOAs) based on blockchain technology and smart contract technology are considered to be the way to realize the autonomy of online communities. However, the existence of algorithmic power has broken this traditional cognition, and metaverse platform companies can quietly control the flow of information and data through the control of the underlying algorithms in the digital environment. With the advantages of algorithm technology, metaverse platform companies can predict users' preferences through in-depth mining of metaverse user behavior information data, and guide them appropriately to the direction of their own will. This is completely different from the centralization that relies on monopoly information sources in the traditional Internet, and is essentially a centralized process that relies on "algorithmic black box" to achieve re-centralization on the basis of decentralization. That is to say, "algorithmic power has a tendency to centralize the power of the platform, and through the dominance mode of implicit control, it has shaped a process that seems to be 'decentralized' but is actually 'recentralized'". The process of information and data decentralization has given metaverse platform companies arbitrary "private power", which is not easily perceived by metaverse users. With this information advantage, metaverse platform enterprises can become the main body of power on the metaverse platform, hiding behind the scenes to maximize their own interests.

Second, the monopoly of economic activities. The metaverse will be the main place for future social and economic activities, and even with the blessing of non-fungible tokens (NFTs), a relatively independent economic transaction system can be built within the metaverse. Based on the decentralization of information and data, metaverse platform companies can achieve a monopoly on economic activities in the metaverse. At present, it is mainly various platform companies that have invested a lot of capital to build, maintain and control the metaverse, and it is these platform companies that can truly control the power of algorithms, that is, "behind the power of algorithms is the power of capital". Profit-seeking is the main driving force for platform companies to promote the development of the metaverse, and the cost investment should be for profit, so platform companies will inevitably use the power of algorithms to maximize their own economic interests. Monopoly is the best way to obtain high profits, so once the platform companies that control the metaverse find that the algorithm has monopolistic power attributes, they will naturally use the power of the algorithm to promote their own interests and build monopoly barriers to economic activities. For example, they can exclude competitors from establishing economic cooperation with users through the metaverse, or they can compile information cocoons that are beneficial to their own interests through algorithms, and secretly place metaverse users in these information cocoons, guide users to make behavioral decisions that are in line with the interests of platform enterprises, and exclude users from being exposed to information that is unfavorable to platform enterprises, so as to maximize the benefits of platform enterprises themselves. This monopoly on economic activity in the metaverse has an invisible cloak, and metaverse users do not feel that they are being coercively interfered with, but the behavioral decisions they make will always be in the best interests of metaverse platform companies.

Third, the manipulation of social concepts. Also based on the recentralization of information and data, metaverse platform companies can realize the guidance and manipulation of social concepts, and even lead social concepts to extremes according to the needs of their own interests. Specifically, metaverse platform companies can set up the digital environment in the metaverse through algorithms, guide the behavior of metaverse users, and indirectly affect users' cognition and behavior in the real world through the "Proteus effect". For example, in the game metaverse, by setting many levels for player characters that require theft, violence, and murder, players can subtly influence their attitudes towards these antisocial behaviors, unconsciously adjust their own behavior in reality to adapt to the identity of the characters in the game metaverse, and even change their own moral cognitive standards in the real world. If this ability of metaverse platform companies is applied to the field of political awareness, they can gradually guide users to their conceptual cognition in the real world by shaping the digital environment of users in the metaverse, controlling the content of information and data they are exposed to, slowly changing their ideology, and even forming extreme political concepts. This potential risk already exists even in the traditional Internet, and the possibility of such risks in the metaverse, which is an advanced form of the Internet, is further increased.

4.2The algorithmic regulation of the administrative regulation organs

Metaverse platform enterprises are the main body of algorithmic power in the metaverse, and the algorithmic power they hold is a double-edged sword, which has both positive significance and the risk of alienation, and must be effectively regulated to prevent the risk of abuse of algorithmic power. It is a feasible way to use the public power of the state to restrain the algorithmic power of metaverse platform enterprises, and the power of administrative regulation is particularly important. Administrative regulation can constrain the algorithmic power in the metaverse through the following paths, reduce the possibility of alienation of algorithmic power, and guide the algorithmic power for good:

First, transparent regulation of algorithms. Algorithmic power essentially stems from the high degree of information asymmetry generated by algorithmic black boxes, and metaverse platform companies have dominating power over metaverse users by virtue of the information advantages generated by algorithmic black boxes. In order to constrain the power of algorithms, the most important regulatory path is to improve the transparency of algorithms. Algorithmic transparency refers to "enabling a neutral third party to detect and review whether there is an algorithmic nuisance by providing information that can be monitored, checked, or reviewed by providing detailed documentation, technically appropriate API interfaces, and permissible terms of use." "The specific regulatory methods include different forms such as the obligation to inform, the reporting of parameters to the competent authority, the disclosure of parameters to the public, the archiving of data and the disclosure of source code. Of course, algorithmic transparency does not mean the complete disclosure of the technical details of the algorithm, and it is also necessary to consider the protection of the intellectual property rights and trade secrets of algorithm designers and users (such as metaverse platform enterprises), otherwise it may curb the innovation momentum in the economic field and even have a negative impact on national interests. Therefore, there must be limits to the transparency of algorithms, and different regulatory paths need to be determined according to specific scenarios. We can distinguish the regulation of algorithmic transparency in the metaverse into two aspects:

First, the algorithm is transparent for the regulatory side of the metaverse. The focus of this algorithm transparency regulation is on the "explainable" of algorithms relative to supervision, focusing on the objective attribute requirements of the technical framework of the algorithm model itself, and emphasizing that the design structure of the algorithm model of the metaverse should enable the professional law enforcement personnel of the administrative regulatory authorities to objectively implement effective supervision of the algorithm model, and does not require the design users of the algorithm to make actual algorithm interpretations. That is to say, "the interpretability of the algorithm itself does not require the actual interpretation of the algorithm, but only requires that the specific mechanism of the algorithm considering various elements can be clearly recognized by professionals and expressed through clear regulatory grammar, which may not necessarily improve the understanding of algorithms by ordinary people, but is very conducive to the development of professional supervision". This kind of algorithm transparency regulation is essentially a constraint on the implementation path of algorithm technology in the metaverse, requiring that the selection of algorithm technology path must meet the requirements of supervision and achieve transparency to administrative regulatory authorities.

Second, the algorithm for the user side of the metaverse is transparent. The focus of this algorithm transparency regulation is to protect the right of Metaverse users to know about algorithmic decisions that affect their rights, focusing on the subjective attribute requirements of the algorithm operation process, emphasizing that Metaverse users whose rights are affected by algorithmic decisions can have the opportunity to obtain explanations and explanations from algorithm design users, understand the setting of key parameters that determine the output results of algorithms, and then obtain stable decision-making expectations. This kind of algorithm transparency for metaverse users can be subdivided into two implementation paths: "algorithm knowledge" and "algorithm selection". The former refers to the requirement that the program design of the metaverse platform clearly prompt that the information that the user is exposed to is the result of the recommendation of a specific algorithm, and clearly inform the user of the algorithm parameters on which the information is pushed, so that the user knows the objective existence of the algorithm's automatic decision-making and related parameters. The latter refers to the requirement that relevant digital options should be set in the metaverse program, so that users can independently decide whether to accept the information automatically pushed by the algorithm, or independently decide what parameters to accept the information pushed by the algorithm, so as to fully respect the autonomy of metaverse users to obtain relevant information and data, and realize the algorithm transparency to the user.

Second, the ethical regulation of algorithms. If the transparent regulation of algorithms is to constrain the power of algorithms by restricting the external form of algorithmic technology, then the ethical regulation of algorithms is to constrain the power of algorithms through the guidance of the intrinsic value of algorithmic technology. The power of algorithms in the metaverse has a variety of potential ethical crises for general artificial intelligence algorithms, including: first, users in the metaverse are treated as objects of algorithms, becoming objects to be calculated and digitized, which actually leads to users losing their human subjectivity in the metaverse and becoming tools for producing information and data; Second, the algorithm in the metaverse takes efficiency rather than the protection of the rights of the subject as the main goal to pursue, and will not hesitate to sacrifice the rights and interests of users in order to achieve efficiency, resulting in the replacement of the value of rights by technical value; Third, the output results of the algorithm in the metaverse depend on the input behavioral information data, which will produce the phenomenon of "garbage in, garbage out", that is, the bias and discrimination contained in the input data will be integrated into the output result of the algorithm after it is run, and present a circular self-reinforcing algorithm discrimination, which will lead to the labeling and stigmatization of specific groups of users, and deny the possibility of self-improvement. It makes it impossible to get rid of the shackles of stereotypes.

Although the potential ethical crisis of algorithmic power in the metaverse is related to the self-learning and evolutionary ability of artificial intelligence algorithms based on big data, it is fundamentally derived from the values of algorithm design users, especially from the values of metaverse platform enterprises that actually control the power of algorithms. Therefore, the regulation of algorithm ethics requires that administrative regulatory authorities should start from the initial source of algorithm design, intervene in the ethical review of algorithm technology in the stage of algorithm research and development and construction, and require metaverse algorithm programs to abide by the value norms of "common good" such as equality and justice in the initial programming stage, integrate these value norms of "common good" into algorithm design, build an algorithm technology framework with a sense of social responsibility, and prevent bias and discrimination in algorithm technology at the source.

Third, algorithmic accountability regulation. Whether it is algorithm transparency regulation or algorithm ethics regulation, the focus of regulation is more on preventing the risk of alienation of algorithmic power in advance. If this preventive regulation does not achieve the desired effect, and the risk of abuse of algorithmic power in the metaverse still occurs, then the allocation of legal liability after the fact is essential. The allocation rule of legal liability is "a complex incentive system", through which the cost-benefit structure of relevant entities can be changed, so as to produce behavioral guidance for them. The allocation rules of legal liability are also an important condition for the implementation of administrative regulations, which provide a legal basis for the administrative regulation of the abuse of algorithmic power in the metaverse. Therefore, in order to realize the goodness of algorithms, it is indispensable to allocate the legal responsibility of algorithms.

In the traditional Internet era, the research on algorithm liability mostly focuses on civil liability, and often defines algorithm liability as product liability based on algorithms, for example, the legal liability of autonomous vehicles after accidents caused by chip algorithm problems is classified into the category of product liability. However, the metaverse is the next generation of the true Internet, and the responsibilities arising from the alienation of algorithmic power are different. This is because the metaverse is not a single product like a self-driving car, but a cross-regional and cross-domain network platform that connects everything, and a virtual world that is intertwined and inseparable from the real world. In this virtual world, people engage in almost the same social and economic activities as the real world through avatars, forming almost the same social interaction relationship as the real world. Therefore, the metaverse already has the form of a public domain, and may even have the form of a "state-like". Therefore, the abuse of algorithmic power in the metaverse also has the characteristics of infringing on the public interest, and should not only be restrained by civil legal liability, but should also be implemented by the administrative regulatory authority representing the public interest by setting up corresponding administrative legal liability, so as to restrict the exercise of algorithmic power in the metaverse. This is the theoretical basis for us to set up administrative legal responsibility for algorithmic accountability regulation in the metaverse.

So in the accountability regulation of the metaverse algorithm, who should be the subject of the corresponding administrative legal responsibility? Some scholars have pointed out that algorithms involve a complex technological ecology, and their development and application will be related to multiple participants, so the corresponding algorithm responsibilities are also fluid and ambiguous, and it is often difficult to define which subject's fault is when producing erroneous results. Therefore, it is not appropriate to adopt the model of multiple responsible entities based on fault distinction to determine the administrative legal liability entity that bears the infringement of algorithmic power in the metaverse, but should adopt a single responsible entity model, that is, the metaverse platform enterprise should be the subject of administrative legal liability for the infringement of algorithmic power. Although the design and application of the Metaverse algorithm will involve multiple subjects, the Metaverse platform enterprise is the main controller of the algorithmic power and the ultimate beneficiary of the operation of the algorithmic power, and only it has the ability to make full use of the algorithmic power to seek its own best interests. Only by allowing the beneficiaries of algorithmic power to bear the cost of the corresponding infringement of algorithmic power can their behavior be truly constrained. Allocating the administrative legal liability for the infringement of algorithmic power to the metaverse platform enterprise can not only incentivize it to restrain the algorithmic power it controls, but also incentivize it to restrain other entities related to algorithms to ensure that algorithms are good. At the same time, the singleness of the administrative legal liability for the infringement of algorithmic power can also enable the administrative regulatory authorities to quickly locate the responsible subject, efficiently implement algorithmic accountability, and improve the efficiency of administrative regulation when the problem of algorithmic power infringement in the metaverse occurs, so as to produce real and effective constraints on the algorithmic power in the metaverse, prevent the alienation of algorithmic power, and guide algorithms for good.

5.Concluding remarks

The metaverse will become the main place for people to conduct social and economic interactions in the future. At present, in the metaverse, which is mainly driven by market players, there may be risks of reducing the possibility of cooperation caused by identity anonymization, individual information and privacy infringement caused by data agglomeration, declining innovation momentum of market entities and national data security risks, and information and data recentralization, monopolization of economic activities, and social concept manipulation caused by the power of algorithms. It is difficult to prevent and resolve these risks by relying solely on the strength of market players, and even the profit-seeking tendency of market players will increase the probability of these risks, making market forces themselves the cause of these risks. Therefore, it is necessary to prevent and resolve these possible metaverse-related risks through the power of administrative regulations. Administrative regulatory authorities can implement mandatory data desensitization, restrict intra-industry competition among metaverse platform enterprises, regulate cross-border data flows, and strengthen the path of algorithm transparency, ethics, and accountability to control these metaverse-related risks, so as to truly become a full-fledged Internet in the new era that enhances overall social well-being.