https://doi.org/10.32575/ppb.2024.1.5
This paper discusses the recently emerging platform law from a jurisprudential point of view. After defining the platform as a general coordination mechanism, it deals with topics such as the rationale for regulation, its main goals, and its general characteristics. According to the study, the main argument for regulation is that the platform, as a coordination mechanism, tends to become unstable without intervention, or to become harmful from the point of view of society. Above all, it tends to abuse the asymmetric power situation that exists between the platform and its users. These conditions must be prevented from occurring, and platform users must be protected in certain situations. The study lists four features that characterise platform law: its ex ante regulatory nature, the predominance of technology regulation and self-regulation, and the extensive use of user protection tools, such as complaint mechanisms, protection of user accounts, and explainability obligations. This toolbox partly resembles the long-established methods of consumer protection, but it also differs from it in certain ways.
Keywords: online platforms, legal theory, theory of code, Digital Services Act, user protection, coordination mechanisms
A cikk a közelmúltban kialakuló platformjogot jogelméleti szempontból tárgyalja. A platform általános koordinációs mechanizmusként való meghatározása után olyan témákkal foglalkozik, mint a szabályozás logikája, fő céljai, általános jellemzői. A szabályozás fő indokaként a tanulmány azt állítja, hogy a platform, mint koordinációs mechanizmus beavatkozás nélkül hajlamos instabillá válni, illetve a társadalom szempontjából káros állapotba kerülni. Mindenekelőtt hajlamos visszaélni a platform és felhasználói között fennálló aszimmetrikus hatalmi helyzettel. Ezeket az állapotokat meg kell akadályozni, és bizonyos helyzetekben meg kell védeni a felhasználókat. A tanulmány négy olyan jellemzőt sorol fel, amelyek a platformjogot jellemzik: az ex ante szabályozási jelleg, a technológiai szabályozás és önszabályozás túlsúlya, valamint a felhasználóvédelmi eszközök, így a panaszmechanizmusok, a felhasználói fiókok védelme és a megmagyarázhatósági kötelezettségek kiterjedt alkalmazása. Ez utóbbi eszköztár részben hasonlít a fogyasztóvédelem régóta ismert módszereire, de bizonyos ponton el is tér attól.
This paper aims to examine the characteristics of the emerging platform regulation from a legal theoretical (jurisprudential) point of view.
Internet platforms became widely used in the wake of the 2008 crisis and soon became the main protagonists of the Internet. However, until the mid-2010s, the legal regulation of platforms was scarce. Most platforms were regarded as an "internet intermediary" (and within that a "hosting" service), which has no direct liability for the activity performed
- 101/102 -
on it.[1] By the end of the decade, it had become clear that the great economic and social significance of platforms demanded a different, and in many respects, much more detailed and strict regulation.[2]
A body of rules to address this regulatory need was created very quickly by the European Union[3] in the period between 2018 and 2022. On the one hand, the EU amended a series of directives to align them with the new phenomenon of platforms (AVMSD,[4] Copyright Directive[5]), and then created new ones (mainly in the form of Regulations) specifically aimed at platforms (P2B,[6] DSA,[7] DMA,[8] Platform work directive draft[9]). The norms reveal a picture of a new European platform law, which has some well-described characteristics. This paper aims to analyse these characteristics on a higher abstraction level, with theoretical ambitions.
The structure of the study is as follows. In the first, methodological part, I clarify some theoretical starting points, such as the rationale, purpose, and justifying principles of platform law and the place of platform law within the legal system. The second part analyses platform law in terms of four characteristics. Three of these are typical of any legal field which deals with technology. These are: the ex ante nature of the law, the extensive use of self- and co-regulation, and the fact that rules are enforced with the help of technology ("regulation through technology"). The fourth, longest subsection will attempt to characterise a concept that is unique to platform law, because it is its main justifying principle. This is the user protection.
- 102/103 -
This paper seeks to characterise platform law on a more abstract level. Its main purpose is therefore not to summarise the regulations currently in force, but to outline the rationale and connections behind the rules, identify their common elements, to organise the rules into meaningful groups, and to demonstrate the organising logic, goals, and justifying principles behind them. It is also not my aim to write primarily about platform theories. This paper is about the theory of platform law and not about the theory of platforms - though I use one of these theories to conceptualise the platform-phenomenon. Another disclaimer is that this legal theory does not operate at the same level of abstraction of the classical jurisprudential tradition (which deals, for example with issues such as the problem of normativity, the values behind the law, the concept of law, or the relationship between law and morality). I have thus chosen a lower level of abstraction for the analysis.
At the same time, despite the fact that it is not my intention to develop a "grand" theory here, the task is still not straightforward. Most of the rules that are specifically tailored to platforms were created only in the last few years, and therefore have no case law to interpret them, along with their so-called ex ante nature. Both characteristics make the task of developing a theory more difficult, because they weaken the organising role of judicial argumentations, which is the usual starting point for doctrinal and jurisprudential explanations.
At this point, it is worth explaining the difficulty caused by ex ante regulation. Since ex ante type (or compliance-type, preventive) regulation was not created primarily as a "decision norm", i.e. as "lawyer's law" (Entscheidungsnormen or Juristenrecht[10]), with the intention of being used by the courts, but instead primarily to prevent problems by regulating the actions of the participants or of the technology developed by them, its operation in many cases remains invisible. Indeed, the very fact that everything is unproblematic and it functions without difficulties, adds little to the exploration of the nature of a rule, because this is more apparent in borderline situations, or in complex cases. Furthermore, ex ante regulation is typically considered by the courts only if someone has violated an obligation, so the legal dispute immediately turns into a question of (ex post) liability. In such cases, the court examines compliance with the ex ante law, and may even come to doctrinal, systematising conclusions, but very often in such situations, the courts
- 103/104 -
appoint an expert witness to interpret the ex ante regulation. In short: from a theoretical and dogmatic point of view, ex ante law is much more difficult to grasp than "lawyer's law".
What triggers the need for the regulation of platforms, and what causes the crisis of the "old" law? There is both a more general and a more specific answer to this.
To understand the general answer, some simple theoretical assumptions must first be clarified. In this study, I regard the platform - following the theory of János Kornai[11] - as a coordination mechanism. The coordination mechanism is an impersonal order that allocates resources and channels human activities. In addition to some less important types, Kornai identified three main types of mechanism: bureaucracy, the market, and ethical coordination.[12] While the market and bureaucracy are impersonal in nature, ethical coordination is deeply embedded in the everyday life of people. The main drivers of market coordination are money and profit, and it channels information through prices. The functioning of bureaucracy is based on explicit rules, (very often legal rules) or direct orders. Although ethical coordination is also aided by rules (ethical rules), it is largely based on human emotions and expectations such as recognition, love, respect, and the feeling of belonging to a community.
The functioning of platforms, as a coordination mechanism, differ from both of these main mechanisms, although they have certain characteristics that make them similar to the impersonal types (market and bureaucracy). The most important feature of the platform is that it can successfully replace the other three main mechanisms, by emulating them. The two impersonal mechanisms, without intervention, tend to become socially harmful, or inherently unstable. Platforms also tend (without continuous corrections) to end up in a dysfunctional state (not fulfilling their original social function, or acting directly against it). Markets have a propensity towards the formation of monopoly markets and cyclical crises,[13] while bureaucracies end to operate with their own interests in mind, which manifests itself partly in expansion and partly in abuses of power and arbitrariness.[14] Platforms build up a special type of inequality, known as "platform power", which makes the user of the platform extremely vulnerable. The relationship between ethical coordination ("the life-world") and platforms is much more subtle. In this regard, a platform transforms realities to data and algorithms that otherwise cannot be quantified,
- 104/105 -
such as love, attachment, or recognition. In this manner, on the one hand, the platform creates a dependency in its users, and on the other hand, since it monetises these goods, it has an interest in maintaining this dependency.
Sophisticated techniques were developed in the past century to eliminate the instability and dysfunctions of "traditional" coordination mechanisms. The correction of the markets is performed partly by economic, (anti-cyclical economic policy) and partly by legal methods (anti-monopoly, antitrust and consumer protection legislation and authorities). For example, in the last century, competition law created a complex system for dealing with monopolies. Consumer protection law also developed a series of well-proven rules in the last 50 years or so to compensate for the asymmetric situation of the consumer and to curb the power of large companies. Rules such as the expectation of comprehensibility of boilerplate contracts, or the prohibition of unilaterally beneficial conditions for the service provider are examples of such law. The coordination mechanism of the market was thus curbed, not by a single set of rules, but by a combination of antitrust and consumer protection rules. The former aims at the "ideal" and "equilibrium" state of the market, while the logic of the latter is more similar to constitutional and administrative law, which seeks to correct abuses of power. It is not by chance that a lively discussion continues amongst specialists in constitutional law about the "horizontal effect of constitutional law",[15] (a somewhat similar issue in the U.S. is known as "state action"[16]) which means the use of certain constitutional principles and fundamental rights in the sphere of private relations between two, legally equal non-state actors.
Constitutional law and, to some extent, administrative law have become the main means of correcting the dysfunctionalities of bureaucracy. Constitutional law and its international counterpart, international human rights conventions humanise political power and the bureaucratic sphere[17] by placing them under the rule of law. The related guarantee rules, from the separation of powers to the obligation to provide reasons for individual decisions, expand and enforce this ideal in certain sub-areas. The entire corpus is interwoven with the means of protecting fundamental rights, which protect the freedom and decision-making autonomy of individuals, and reduce their vulnerability.
Since the platform is a new coordination mechanism in the virtual space, it can be present in almost all spheres of life and can replace bureaucracy, the market, or even "ethical coordination" - the normal "offline" private interactions. At the same time, control mechanisms similar to those developed to counteract the dysfunctionalities of the various coordination mechanisms in these fields are not in place for platforms: the platform as a coordination mechanism has not yet been subjected to these rules. The new platform law, which will be discussed here, is essentially an effort by the legislators now trying to create these mechanisms, based on the closest analogues: constitutional law, data
- 105/106 -
protection law, and consumer and (possibly investor) protection rules.[18] These rules were codified by the European legislator in line with the current fashion for ex ante rules with their own agency and sanctions, which make extensive use of self- and co-regulation tools, along with technology-based regulation (using the technology-oriented and technology-regulated method).
The other question that should be explored before discussing the individual characteristics of platform, is its place in the legal system as a whole, within the legal corpus.
Platform law is always Janus-faced, because it not only has to deal with effects arising from the characteristics of platforms such as datafication, algorithmic control, and network effects, but must also be adapted to the specific areas of life in which the given platform functions. A work platform must take into account the specific risks arising in the world of work (e.g. compliance with rest periods) together with the effects arising from the fact that it is basically a marketplace platform connecting supply and demand. An accommodation platform must also take into account the characteristics of the field of tourism and accommodation rental. At the same time, sectoral logics never completely dominate the logic of platforms: it is possible to refer to platform law precisely because the effects arising from the characteristics of platforms are so strong that they prevail in all areas. Certain approaches (especially certain means of "user protection") permeate all facets of platform law. Of course, this duality often appears in practice in the reverse order: the regulators start from the logic of the given sector (e.g. from the logic of media law or of labour law in the case of the AVMSD or the Work-Platform Directive), but are usually subsequently also forced to take into account the logic of platforms.
This duality produces a situation where the law very often uses long-established approaches in a given field to regulate platforms operating that sector, and some of these well-established rules can of course be effective. The advent of platforms does not make it necessary to set aside all the old legislative goals (or other justifying principles). The AVMSD is an excellent example of how the goal of protecting minors, for example, should also apply to video-sharing platforms. The other issue is whether specific measures will have a different effect on a platform. In linear television, the placement of content harmful to children in late-night broadcast slots, for example, has proven to be quite effective for several decades, while in the case of on-demand content, this approach is so far not applicable (this is not yet a platform feature) to the recommendation algorithms operating on the platforms and their responses to user behaviour, although the risks of using data require specific, platform-legal solutions.
- 106/107 -
In platform law, three basic nodes or sub-areas are emerging in the regulation: on the one hand, there is the "general platform law", which mainly includes the definition of platforms and issues related to their responsibility and transparency. "Specialised" platform law is again divided into two parts, one of which is closely related to fundamental rights, and the other more infused with economic logic. The former views platforms from the perspective of online dangers and the risks of violations of fundamental rights.[19] Competition law, copyright law, consumer protection and labour law rules can be grouped around the economic node. These are more similar to consumer protection and antitrust rules. (The Copyright Directive is an exception to this, where the protection of the rights of authors and publishers is dominant.)
The regulations of the EU and the USA may be very similar at the level of individual detailed provisions and certifications, but they differ greatly in their overall approach. First of all, the U.S. did not create a "horizontal" or "general" platform law, as the EU does with the DSA.[20] The EU's general platform law, and the effort to address the platform phenomenon in general, is probably based on the comprehensive aspirations of European codification traditions. At the same time, generalisation also results in the fact that there must be a "special part" related to it, reflecting the Janus face described above. It must fit both the concepts and the regulatory environment specific to the sector and the "general platform law".
More importantly, however, the justifying principles and arguments applied in the USA differ in general and in individual sub-areas. U.S. law, for example, is more hesitant when it comes to the "precautionary principle" than Europe, where this is one of the most important justifying principles, [21] and one to which I will return. Some examples of the differences in sub-areas include: the topic of "illegal content" in the EU, is most often referred to as the topic of the "first amendment", i.e. freedom of speech, in the U.S. What is a data protection issue in the EU, in the USA either simply does not exist as a problem, or it is listed as a "privacy" issue that only partially overlaps with data protection. Competition law problems are often mentioned as "common carrier" issues, which is also a specifically American concern, unknown in the EU. Finally, comparing the situation of the USA and the EU is made very difficult by the fact that the legislation of the USA is fragmented both horizontally and vertically, and is very often "situational" legislation. Beyond vertical
- 107/108 -
(federal and state) segmentation, horizontal fragmentation means that, for example, the federal legislation reflects on certain partial problems and does not seek to deal with the issue of platforms as a whole. A good example of this is the ACCESS Act, which aims to solve the problem of data portability[22] (which in Europe is not part of platform regulation at all, but of data protection law[23]).
Taking all of this into account, I characterise platform law with reference to four concepts or pairs of concepts that have sparked legal discourses in recent decades in other fields. What they have in common is that they try to describe and capture the set of rules that is often contrasted with "law", which is the "matter" of lawyers, as "regulation".
Among these, the first three concepts characterise not only platform law, but also the rules created for the technology-embedded world in general, and EU legislation in particular.[24] These aspects are: the predominance of ex ante regulation, the intensive reliance on co- and self-regulation, as well as the increasing role of technology regulation and regulation with the help of technology. These characteristics are therefore those that characterise modern legislation anyway, but they are of outstanding importance in platform law too.
The fourth feature is a feature of platform law that is unique to it. This is the justification and purpose of platform law: user protection.
One of the key features of platform law is that it is largely ex ante in nature, or in other words, compliance regulation.[25] This characteristic has been brought up many times in this book, but here I would like to discuss it on a slightly more abstract level.
The mass appearance of these rules began in the 1960s and 1970s. The paradigmatic case of ex ante law is environmental protection and pharmaceutical regulation, in which the justification of ex ante rules also received a special name: the precautionary principle,[26] a kind of legal version of the folk wisdom "better safe than sorry".[27]
The ex ante - ex post distinction itself appeared mainly within the discourse of law and economics starting in the 1990s, usually in the context of how far it is reasonable to go in ex ante regulation, (in contrast with the ex post - liability - rules), and what the
- 108/109 -
advantages and disadvantages of both forms are and how can they complement each other in a good regulatory mix.[28]
Since the 1970s, it has therefore become increasingly accepted that the purpose of the law is not to respond to some kind of illegal activity, but to facilitate the avoidance of certain risks, by providing participants in a given risky area with behavioural or direct technological specifications. This right is often called regulation, or compliance, in order to contrast it with law.
A very straightforward and easily understandable example of the distinction between ex post and ex ante rules is the system of norms regulating traffic on public highways. In this field, on the one hand, there are the traffic rules, which is a classic ex ante regulation, as they coordinate road users with a series of specific regulations, and their main goal is to prevent accidents. However, traffic is also regulated by ex post rules: liability for accidents in the civil law, or traffic offences in the penal law. Further notable differences exist between the two systems of rules, such as the fact that traffic rules must be constantly followed, while the liability rules only come into play in the event of an accident. Moreover, traffic rules are constantly applied and interpreted by civilians, while the rules of responsibility are primarily interpreted by lawyers, and within this, especially by the courts.
Ex ante law is not homogeneous, and this also characterises platform law. Some rules regulate processes, assess and manage risks, and possibly contain certain quality assurance elements. This can be characterised as risk-preventing ex ante law. On the other hand, other rules deal with inequalities, injustices, and power asymmetries, very often in a consumer protection setting. While the key concept of the latter field of law is "power" and its correlative "vulnerability", the focus of the former is "risk", "health", "security". Even more simply, some ex ante law protects a person's health, and some protects one's dignity and other values arising from it (e.g. decision-making autonomy, freedom of speech, etc.). The majority of the ex ante rules of platform law naturally fall into the latter category.
How the centre of gravity of the law has shifted towards ex ante rules is best demonstrated by the change in one of the central constructs of platform law, the notice and takedown rule. The notice and takedown rule described in the E-commerce directive was clearly an ex post rule: it determined when a platform is not responsible for violating content and when it is.[29] The obligation of the platform (hosting service) clearly began after the publication of the illegal content. Compared to this, the first sentence of the new notice-and-action regulation
- 109/110 -
stipulates that the providers of hosting services "put mechanisms in place to allow any individual or entity to notify them of the presence on their service of specific items of information that the individual or entity considers to be illegal content".[30]
It is important to add that the distinction between ex ante and ex post rules is relative in many respects. Some also argue that the ex ante function of criminal law (to prevent people from committing crimes) is actually more important than its ex post function (to punish those who commit crimes).[31] Based on this argument, the distinction between ex ante and ex post regulation actually makes no sense, since one of the most important goals of all ex post rules is to avoid problems, and the (secondary) goal of ex ante regulation is also that in the event of a problem occurring, it can help guide in the determination of responsibility. At the same time, this argument is flawed, in that the ex ante and ex post distinction does not (only) apply to the purpose of the law, but also to the method of regulation. Ex ante traffic rules describe an "ideally safe", "accident-free" flow of traffic, and because of this, it is no longer up to the driver to decide how to drive safely. Henceforth, he can be punished even if he simply breaks the traffic rules; this no longer requires the occurrence of a specific accident, or injury, or even a dangerous situation arising. This phenomenon, whereby the law pre-empts other (moral or practical) arguments is one of the leitmotifs of modern legal positivism.[32] However, as will be discussed below, this ex ante nature, together with the embedding of rules in technology, creates a new quality that limits rather than promotes the unfolding of human freedom.
The relationship between the two regulatory methods has been accompanied by heated debates, on points such as which type of regulation is more effective and how far it is reasonable to extend the generally heavy administrative burdens that are typically associated with regulation, such as registration and reporting obligations, the operation of separate monitoring systems, employing officials, and setting up regulatory and enforcement authorities.[33] In this regard, the (not very telling) argument is acceptable that beyond a certain point, ex ante regulation is counterproductive and it should therefore not be exaggerated, because the transaction costs will be too high, or because it is not worth to deal with risks that have a very low probability of occurrence.[34] The most appropriate form of regulation is a "mixture" of ex ante and ex post regulation. At the same time, other affected parties (including many legislators) argue in favour of strengthening ex ante regulation,[35] and it seems that this position is also more dominant in the field of European platform law. Tracing the textual changes made during the drafting of the DSA, it was clear
- 110/111 -
that it was continuously being enriched with new ex ante elements. It were not primarily completely new obligations that were being added, but rather an increasingly detailed description of the fulfilment of already defined ones.[36]
Platform law, like European law as a whole, makes extensive use of co- and self-regulation. The idea of co- and self-regulation, which partly overlaps with the notion of "soft law",[37] gained ground in EU law at the end of the 1990s. At that time, a debate about the institutional efficiency of the EU began, which ultimately led to the introduction of co-regulation as one of the most important recommendations of the White Paper on European governance.[38] According to the White Paper, the essence of co-regulation is that the EU establishes only the general rules, the details of which must be worked out by the industry participants. The findings of the White Paper were incorporated into the action plan on better regulation.[39]
Platform law makes extensive use of self-regulation, as well as the outsourcing of regulation to private organisations.[40] Tus, for example, according to the text of the DSA, the Commission supports and promotes the development of various industry standards, e.g. for the electronic reporting of (perceived) illegal content, the creation of templates, designs and procedural standards that make it easier for users to understand the restrictions included in contract terms, for an electronic reporting system of reliable whistle-blowers, etc. The same applies to codes of conduct: the Commission wishes to encourage the creation of voluntary codes of conduct. Similar self-regulation provisions can also be found in the AVMSD and other platform law norms.
It may seem that in certain respects the notion of co- and self-regulation runs counter to the trend of ever-increasing ex ante regulation. As noted above, the trend has been towards a growth in the amount of increasingly detailed ex ante rules. At the same time, the detailed description of obligations narrows the room for manoeuvre of legal entities rather than expand it. This seems to be a paradox, at least in platform law, but this is only on the surface. The vast majority of ex ante regulations in platform law are of a formal and procedural nature, as I will detail in the subsection entitled "user protection". At the same time, this also means that the legislation mostly leaves substantial issues to be decided by the platforms. The definition of the specific forms of expression of "speech that can
- 111/112 -
no longer be tolerated" or of the specific forms of practices that are considered detrimental to consumers (to mention only two very important content issues) is essentially completely left to the platforms. From this point of view, I also classify it as a matter of content when the platform rules refer to "appropriate technical measures". In such cases, it is up to the platforms to determine what services they operate in order to achieve a given goal, what functions they implement, and what user interfaces they design for them.
This logic actually suggests that "it doesn't matter what kind of regulation there will be, as long as there is something and it is predictable". Although, we leave its precise content to the profession that previously cultivated it based on its internal rules. However, unlike medicine, driving a car, or the banking profession, platform operation is not an independent profession that had previously established rules which only need to be translated into legal form. The platform is only an "empty" mechanism that coordinates social life, certain markets, etc. In other words, banking law can be outsourced in such a way that prudential obligations are codified as new rules of banking, and within the framework of this, the rest can be entrusted to the bankers, but - I would argue - the same method will not work smoothly if applied to platforms.
One of the most important parts of platform regulation is the direct regulation of technology and/or regulation with the help of technology.[41]
To establish a broader context, it is worth recalling Brownsword's recently published book. In this work, he identifies three ways of discussing the law, or three legal mindsets: law 1.0, 2.0, and 3.0.[42]
The correct application of the rules is at the heart of the Law 1.0 way of thinking. A lawyer's special knowledge concerns how to apply the general rules to individual cases. Such application can be quite easy, a little more complicated and quite difficult. Sometimes difficulties arise due to technological development. For example, such difficulties were caused by the railway, electricity (in private liability), photography (in criminal law) and the rise of tabloid newspapers (in the field of privacy), or the VHS tape recorder (in copyright law). These difficulties may mean that the old legislation has become unusable, or that there are "gaps" in its application.
Law 2.0 is a type of discourse and an approach, according to Brownsword, somewhat similar to what was characterised above as "ex ante" regulation, but with some important additions. The aim is therefore to avoid risks, but this burdens the regulation with a series of new dilemmas and tasks. Finding the optimal point between under-regulation and over-regulation[43] is a new dilemma, and a new task is presented by this type of regulation
- 112/113 -
requiring control and enforcement mechanisms (procedures, authorities). Finding the right balance between ex post liability rules (representing Law 1.0), and new type of ex ante compliance rules (representing Law 2.0) also poses a great dilemma.
Technology plays a different role in the thinking of law 3.0. If the goal is to avoid certain risks, then sometimes these risks can simply be handled better by employing certain technologies. Brownsword cites the 2009 Västberga robbery in Sweden as an example. In this case, a group carried out a helicopter robbery at a cash collection point. Some of the money was never recovered, prompting the Swedes to radically reduce their use of cash. It is easy to see that electronic money cannot be stolen using traditional methods: technology can practically completely filter out a specific type of risk. The same is true, for example, of copyright protection. Although, it is possible to stipulate in the user agreement that it is forbidden to transfer an electronic file to another person, it is much more effective to prevent this by technical means and make the file uncopiable. In fact, in this case, it is a matter of embedding either the ex post (liability rules if the person has copied them) or even the ex ante rules into a specific technology, which in the majority of cases takes the form of some code, software, and (less often) even a physical device.[44]
It is not difficult to discover the similarities between Brownsword's theory and Lessig's theory, as well as the American "lex informatica"[45] discourse, or even Richard Susskind's idea of "embedded legal knowledge".[46] The essence of all these is that legislators use the direct description of technology in order to control the behaviour of people.[47]
In platform law it is extremely common for a technology to be defined to achieve certain legislative goals. Basically, in the DSA, these regulatory (behavioural control) technologies are present in two ways. Either they are provisions where technology is specifically referred to, the DSA is full of such regulations, such as the internal complaint handling system,[48] and the adjustable or non-profiling recommendation system,[49] or they do not refer to them directly, but the given obligation can essentially only be implemented by technological means. (The DSA often uses the term "mechanism" or "necessary technical and organisational measures"[50] in this case as well, which of course can also take the form of a tightly regulated procedure without technological support, but most of the time,
- 113/114 -
these rules involve a specific technological solution. The notice-and-action "mechanism" itself is a good example.[51] It is no coincidence that the preamble of the DSA specifically mentions that the efficient and uniform application of the obligations contained in the Regulation requires technological tools, so it is important to encourage the development of voluntary standards.)[52]
Behaviour control through technology raises countless theoretical and practical problems, which cannot be comprehensively discussed here. Instead, I will outline four of them. First, since the law is most often manifested in the form of language, ordinary people can easily access it if it is formulated in an understandable way.[53] At the same time, understanding the rules wrapped in technology can be difficult. Second, and closely related to this, if we perceive only the outputs of a given regulatory technology, this leads us into the complicated area of algorithmic transparency and explainability. Third, the unclear relationship between legal rules and technological rules also raises questions of legitimacy: who and how are these rules created, and does the political community have a say in the process? Finally, fourthly, technical rules typically have a direct coercive effect on us and cannot be broken, or infringed, which is not always good. Legal rules sometimes need to be broken, not only on the basis of, for example, value considerations, but also for practical reasons.[54]
The three regulatory characteristics mentioned above (ex ante, self-regulation and regulation with technology) may all be present, and may be related to each other. Regulation may thus be both ex ante type, directed at a particular technology (which otherwise complies with a rule), and at the same time, the specific form this regulation takes is entrusted to the self-regulation of the platforms. As I mentioned above, these three characteristics are not only specific to platform law. They characterise, for example, the latest draft legislation related to artificial intelligence[55] or the data economy,[56] which is related to platform law in many ways.
I argued above that the main problem with platforms is the same as that which bedevils the other two large coordination mechanisms: that they tend to become dysfunctional in
- 114/115 -
the absence of intervention. I also indicated that dysfunctional operation can mean several things in the case of platforms.
First, since the role of gatekeepers (news outlets) has ceased or weakened due to "cheap speech" on platforms, users are very easily able to post illegal or harmful content. Second, the abuse of "platform power"[57] can put users in an unjustifiably vulnerable situation. Of course, such "abuse" is sometimes in quotation marks. Often there is no real abuse, - which describes the actual behaviour of a human being towards another human being. It is very often the case, instead, that algorithms "abuse" their "power" on the platform, and even when there is human intervention, it is not directed against someone personally. However, the power imbalance and the vulnerability are still present.
The interaction between the platform as a coordination mechanism and other mechanisms also raises several problems. Sometimes the platform takes the place of other coordination mechanisms, but it is also possible that it plays a complementary and sometimes distorting role. In such cases, the key issue of the regulation of platforms will be how much interest is attached to preserving (conserving) the original functioning of the replaced (emulated, or distorted) mechanism. How important is it for the market to function as it did before the emergence of algorithms? The fact that the customer does not make a decision in an information-deficient environment while only having a very restricted overview of what is on offer is obviously not a value that necessarily has to be protected: the platforms are able to show the customers the entire offer when appropriate, and this is precisely their biggest advantage. This, however, gives rise to a new problem, the question of ranking. In the information-deficient world before platforms, if a consumer wanted to buy a used car, he/she went to the nearest used-car dealer, or bought an advertising newspaper. Now, it is possible to search for the given car type and see the cars for sale in a certain order, defined by algorithms. This abundance of information - it seems - gives birth to a new regulatory need: the need to make the ranking fair and sometimes understandable.
At the same time, the elimination of the information deficit situation is not a positive development in other cases, and intervention may be necessary to maintain it. For example, the seller does not need to have the right to access all the information about the users, such as the consumer's searches before making a purchase decision, or purchase and payment histories. On the seller's side, in some cases, there is more of an interest in preserving the information deficiency.
An even more exciting area is the interaction between the "world of life" (or life world),[58] coordinated by "ethical coordination" and algorithmic coordination. Social media tries to replace and emulate traditional human interactions in many respects, but in other places it instead complements and enhances it. However, the substitution does not work in the original form either. Posting news of a family event (e.g. "my daughter got married") had no equivalent in traditional interpersonal interactions. A person could tell a small group of
- 115/116 -
friends, or could tell one person at a time the good news, but she could not tell 500 people at the same time. Moreover, the question from then on is whether the law has anything to do with this, and whether traditional forms of interaction should be protected.
As discussed above, one of the manifestations of the dysfunctional operation of impersonal coordination mechanisms is that algorithms are capable of "oppression", i.e. creating asymmetric power situations, almost without any human intervention. The law, however, is not set up to deal with this impersonal (algorithmic) power. The law can only deal with asymmetric situations that have been created by people, or by institutions (legal entities), albeit not necessarily always intentionally. Initially, these asymmetric situations occurred, and were handled in relation to political power, and later also in the context of private powers (labour law, consumer protection), and now the platforms have also created the context of impersonal mechanisms ("repressive algorithms").
The problem is twofold: the law is perfectly suited to curbing the exercise of human power and to mitigating human vulnerability, and very sophisticated mechanisms have been developed to address this in recent centuries. The intertwined systems of the legal systems of nation states and of international treaties, along with the sophisticated doctrines surrounding the human rights system, the polished system of institutions and legal protection mechanisms, and the solutions of individual branches of law are part of a huge, well-functioning system. For example, the system of collective labour law in the field of labour law, international human rights conventions and constitutional law and the legal protection activities of international courts, ombudsmen and constitutional courts, as well as consumer protection law, the extensive system of rules related to boilerplate agreements and consumer protection,[59] competition law and sectoral (e.g. financial supervisory) authorities are mechanisms that have been operating for decades.
These legal instruments are not suitable for curbing the power "exercised" indirectly by technical means, and also not suitable for dealing with the dysfunctional operations caused by algorithmic coordination replacing ethical coordination. In the system of constitutional law, the mechanisms that are capable of truly limiting political power and preventing tyranny, from general elections through independent courts to parliamentary motions of no confidence, have been developed over the past 200 years. At the same time, technology has not yet possessed power-related dimensions, or to put it more precisely, technology itself has not been able to "exercise" power independently and autonomously, and thus to make people vulnerable. Technology has always had the capacity to endanger human lives and physical integrity, but it could not make people vulnerable, and above all, it could not, for example, manipulate, change, divert, or provoke large social processes on a mass scale. Although the mass media, for example, has had a huge impact on the public for a long time, and is able to seriously influence people's thinking, actions and mental state, this influence is not wielded by the technology itself, but by the people, editors, and journalists
- 116/117 -
who are in possession of the technology. For this reason, all previous media regulation was naturally aimed at these people (the people behind the media providers, the people hiding behind the mask of the "employer", the "big company"). However, other options present themselves when it comes to technology. The dangerousness of a drone can also be reduced through technological regulation by uploading airspace restriction data to drones before their flights, causing the drone to simply not fly into certain zones.[60]
However, when it comes to regulating platforms (and artificial intelligence), the law is facing something completely new. There is already a debate about whether the technology itself causes or creates vulnerable or powerful situations, and not, for example, the people behind it. This raises the question of whether this new narrative is valid, or is it just an old one appearing in new clothes. The problem is very complex, but the answer may basically be that people and technology are intertwined, and form one system. The operation of the algorithm that makes decisions about workers on a work platform is affected by how it is "tuned" by the operators of that work platform, as well as how the "employers" offering the jobs on it behave, or what qualifications they expect from employees. The same is true for other algorithms. The items that Facebook puts on an individual's news feed depend, of course, on the settings of the news feed, but also on other people's and on the user's behaviour.
There are two main legal sources of inspiration for user protection: consumer protection and investor protection, but platform law introduces many modifications to the legal solutions from these two areas of law. For example, the focus of European consumer protection is unfair general contract terms, which is a kind of blacklist of provisions that should not be used in these contracts.[61] Platform law does not operate with a negative list of this kind, but instead positively lists what must be included in the contracts. Among these mandatory elements, the platform's decision-making powers related to user profiles, and user content play a very important role. Platform law stipulates that the terms and conditions of platforms must include restrictions on the use of their services or complaint handling procedures.[62]
In short: the ultimate goal and justification of the platform law, then - along with some other, equally important, but perhaps subordinate goals, such as preserving the healthy structure of the public sphere or maintaining competition - is primarily the protection of users from the dominance of platforms or sometimes from the harmful and dangerous behaviour of other users. How is this power manifested? It can primarily be found in the way that individual freedom and (decision-making) autonomy may be impaired due to the specific operation of the platforms.[63] Platforms are able to invade an individual's private sphere to an unprecedented extent, learn about their behaviour, collect data about people and transactions, and manipulate users with the help of
- 117/118 -
microtargeting. The situation is made worse by the fact that the monitoring and data collection are mostly carried out by algorithms, i.e. impersonal mechanisms, which even make a series of decisions. This is compounded by the fact that in the meantime, in certain spheres of existence (social public and some market segments), the platforms have gained an overwhelming significance, and have become unavoidable, so that it is very difficult or even impossible to get by without them.
Online safety and the protection of users, especially minors, is constantly emphasised by communication related[64] to the Digital Services Act concerning illegal (and harmful) content. The underlying logic is similar to the corresponding institutions of media law, and in the case of the AVMSD, the rules for electronic media must also be applied to video-sharing platforms in this regard. What makes platforms' obligations regarding illegal content very different from that of the media is the (theoretical) lack of prior screening and general monitoring obligations. The E-commerce Directive only codified the notice-and-takedown procedure in relation to illegal content where the hosting service provider only has to deal with the illegal content ex post if it becomes aware of it. This main rule, in an ex ante form, was also retained by the DSA, but with an extremely large number of exceptions and limitations.
One of the limitations is that the absence of a monitoring obligation does not involve a monitoring ban, and the platforms have monitored the content published on them from the outset.[65] The other is that the various platform law norms and the case-law have established a series of exceptions to the general lack of obligation.
The legal toolbox of user protection consists of five major areas: 1. protection against illegal content, 2. prescribing the mandatory content of user contracts, 3. protection of the user's digital identity, i.e. accounts and digital freedom of speech, 4. transparency of algorithms, and 5. complaint handling rules for operating mechanisms.
Firstly, the sui generis approach of platform law to illegal content is a preventive (ex ante) system consisting of three lines of defence. The first element is the provision and detailed regulation of user-friendly, easily accessible interfaces for reporting illegal content.[66] The second is the system of trusted flaggers.[67] Finally, the third set of rules prescribes protection against abuse.
- 118/119 -
The second means of user protection is the mandatory provision of certain content elements in the contracts concluded with users (or their general terms and conditions).[68] The DSA had already stipulated that intermediary service providers (i.e. a broader category than the platform), must provide information in their contracts about the restrictions introduced in connection with the use of their services, such as details of the content moderation employed, including algorithmic decision-making and human review.[69] Compared to intermediary service providers, platforms have even more serious obligations. For example, in their contracts with user platforms must clearly describe, in detail, what kind of policy they pursue in relation to users who publish blatantly illegal content and who notoriously unreasonably report others.[70] Platforms additionally have to indicate the main parameters used in their recommendation systems,[71] as well as the alternatives that may be available to users of the service to modify or influence such parameters. This must also be included in their contract terms and conditions,[72] so that the transparency of the algorithms is ensured in the contracts.
The minimum requirements for contracts between the platform and the user can also be found in the Platform-to-Business (P2B) regulation. The first set of rules regulates some of the characteristics of the contracts between platforms and entrepreneurs. One such rule requires the clear and comprehensible wording of contracts. The provision appears in almost the same form in Article 5 of Directive 93/13/EEC. The P2B regulation requires that the reasons for a decision to suspend, terminate or otherwise restrict a user's account be indicated as a mandatory element of the contract.[73] The same article concerning contracts includes a provision on how to notify the user of changes to contracts and what grace period is required for them to take effect.
The third group of user protection rules concerns the protection of digital identity and freedom of speech. This curbs and controls the platform's decisions that affect users most deeply (primarily exclusion from the platform, suspension, or restriction).[74] For example, as we have seen, the DSA requires the operation of an efficient and easily accessible internal complaint management system,[75] which can be used in these cases. P2B, in addition to imposing certain formal requirements on these decisions (communication on a "durable medium", 30 days prior notification in the event of termination),[76] imposes an obligation to justify disciplinary measures (in addition to the internal complaint handling mechanism included in the DSA, the platform work directive requires[77] that they be in written form
- 119/120 -
and the possibility of human review of (algorithmic) decisions that result in the restriction, suspension or termination of the profile (account) of a platform worker.[78]
The fourth tool of user protection, contained in the provisions of all three documents, attempts to make the operation of algorithms that affect users more transparent during everyday use. In connection with the giant platforms, the DSA requires the "main parameters used in the recommendation systems"[79] as well as "any options for the recipients of the service to modify or influence those main parameters"[80] to be included in the contract for the users of the service. The other two norms - since the stakes in both areas are much higher than on a social media platform - are likewise much more detailed in terms of ensuring algorithm transparency. The P2B Regulation, which mainly protects (small) businesses operating on large marketplace platforms, devotes a separate article to provisions related to the transparency of "ranking".[81] According to this article, "intermediary service providers" must record in the contract "the main parameters determining ranking and the reasons for the relative importance of those main parameters as opposed to other parameters". In addition, search engine service providers must also disclose the main parameters that play the most significant role individually or together and their relative importance. Moreover, in the platform work directive, a whole separate chapter deals with issues of algorithmic management.[82] This chapter not only includes rules related to transparency and explainability, but also - in a manner which is otherwise exceptional in platform law - certain substantive rules, that is, regarding what the algorithms of work platforms should not be, e.g. they must not put undue pressure on workers or otherwise endanger their physical or mental health. In addition, as mentioned above, in the case of certain algorithmic decisions, a written justification and the possibility of contacting a person must be provided.[83]
The fifth characteristic tool of user protection is the introduction of various dispute resolution and complaint handling mechanisms. As we have seen, this tool is often intertwined with the first two, because it provides the possibility of redress against the most important decisions or decisions made by algorithms, but not always. It seems that the regulations analysed here consider complaint handling mechanisms as general user protection tools. There are two types of such mechanisms, internal and external mechanisms. In the case of external mechanisms, complaint handling or dispute resolution takes place not within the platform, but independently of it.[84] Trusted flaggers can also be regarded as such a mechanism. Internal mechanisms include the complaint handling mechanism of the DSA[85] and the mechanisms regulated in Article 7 of the draft directive
- 120/121 -
of the work platform regulation. The successor of the old notice-removal mechanism, the notification and action mechanism, can also be considered an internal mechanism.
A completely separate area of user protection is the set of rules that primarily prescribe compliance, along with the transparency rules for intermediary services and platforms. This difference is partly related to the fact that the platforms are obliged to continuously disclose their efforts regarding individual user protection, as well as the data related to them. In the DSA, intermediary service providers had already been subject to some transparency reporting obligations, while platforms and VLOPs are subject to even more additional obligations.[86]
The transparency reporting obligation of intermediary service providers mainly involves the communication of information on content moderation. In accordance with this, they must submit an annual report on content that was removed on the basis of external or internal initiatives, according to the type of illegality. Online platforms are obliged to regularly prepare reports on suspensions, matters that were referred to dispute resolution bodies, the functioning of content moderation algorithms, and their number of active users, among other things. Furthermore, the giant online platforms have such a wide set of reporting obligations that there is no space to describe them in full here. By way of illustration only, in addition to the obligations relating to the platforms, they are required to operate a database of online advertisements, provide the Commission with access to essentially all of their data, produce and publish a report on risk assessment and mitigation measures and to undergo and publish the results of independent audits, among other obligations.
Bibó, István (1986): Az európai társadalomfejlődés értelme [The Meaning of European Social Development]. In Válogatott tanulmányok 3. kötet [Selected Studies, Volume 3]. Budapest: Magvető, 7-122.
Boehmer, Christiansen (2002): The Precautionary Principle in Germany - Enabling Government. In O'Riordan, Timothy - Cameron, James (eds.): Interpreting the Precautionary Principle. London: Routledge, 31-61. Online: https://doi.org/10.4324/9781315070490
Brownsword, Roger (2005): Code, Control, and Choice: Why East is East and West is West. Legal Studies, 25(1), 1-21. Online: https://doi.org/10.1111/j.1748-121X.2005.tb00268.x
Brownsword, Roger (2011): Lost in Translation: Legality, Regulatory Margins, and Technological Management. Berkeley Technology Law Journal, 26(3), 1321-1366.
- 121/122 -
Brownsword, Roger (2021): Law 3.0. London: Routledge.
Brownsword, Roger - Scotford, Eloise - Yeung, Karen (2016): Law, Regulation, and Technology: The Field, Frame, and Focal Questions. In Brownsword, Roger - Scotford, Eloise - Yeung, Karen (eds.): The Oxford Handbook of Law, Regulation, and Technology. Oxford: Oxford University Press, 3-38.
Cauffman, Caroline - Goanta, Catalina (2021): A New Order: The Digital Services Act and Consumer Protection. European Journal of Risk Regulation, 12(4), 758-774. Online: https://doi.org/10.1017/err.202L8
Cohen, Julie (2017): Between Truth and Power. The Legal Constructions of Informational Capitalism. New York: Oxford University Press.
Darley, John M. - Robinson, Paul H. - Carlsmith, Kevin M. (2001): The Ex Ante Function of the Criminal Law. Law & Society Review, 35(1), 165-190. Online: https://doi.org/10.2307/3185389
Dumbrava, Costica (2021): Key Social Media Risks to Democracy: Risks from Surveillance, Personalisation, Disinformation, Moderation and Microtargeting. European Parliamentary Research Service. Online: https://doi.org/10.2861/135170
Ehrlich, Eugen (2022): Grundlegung des Soziologie des Rechts (5th edition). Berlin: Duncker & Humblot. European Commission (2001): European Governance: A White Paper. COM (2001) 428. Online: https://ec.europa.eu/commission/presscorner/detail/en/DOC_01_10
European Commission (2002): Communication from the Commission of 5 June 2002, Action Plan "Simplifying and improving the regulatory environment". COM (2002) 278 final. Online: https://eur-lex.europa.eu/EN/legal-content/summary/action-plan-for-better-regulation.html
European Commission (2020): Proposal Regulation of the European Parliament and of the Council on European Data Governance (Data Governance Regulation). COM (2020) 767 final. Online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A52020PC0767
European Commission (2022): Proposal for a Regulation of the European Parliament and of the Council on Harmonised Rules on Fair Access to and Use of Data (Data Sharing Legislation). COM (2022) 68 final. Online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM%3A2022%3A68%3AFIN
Fried, Barbara H. (2003): Ex Ante/Ex Post. Journal of Contemporary Legal Issues, 13(1), 123-160. Online: https://doi.org/10.2139/ssrn.390462
Galle, Brian (2015): In Praise of Ex Ante Regulation. Vanderbilt Law Review, 68(6), 1715-1759.
Gardbaum, Stephen (2003): The Horizontal Effect of Constitutional Rights. Michigan Law Review, 102(3), 387-459. Online: https://doi.org/10.2307/3595366
Habermas, Jürgen (1987) [1981]: Theory of Communicative Action, Volume Two: Lifeworld and System: A Critique of Functionalist Reason. Translated by Thomas A. McCarthy. Boston, Mass.: Beacon Press.
- 122/123 -
Harremoës, Poul - Gee, David - MacGarvin, Malcolm - Stirling, Andy - Keys, Jane - Wynne, Brian - Vaz, Sofia Guedes eds. (2001): Late Lessons from Early Warnings. The Precautionary Principle 1896-2000. Luxembourg: Office for Official Publications of the European Communities.
Hildebrandt, Mireille (2018): Primitives of Legal Protection in the Era of Data-Driven Platforms. Online: https://doi.org/10.2139/ssrn.3140594
Joh, Elizabeth E. (2016-2017): Policing Police Robots. UCLA Law Review Discourse, 64, 516-543.
Keynes, John M. (1936): The General Theory of Employment, Interest and Money. San Diego - London - New York: First Harvest/Harcourt.
Kolstadt, Charles D. - Ulen, Thomas S. - Johnson, Gary V. (1990): Ex Post Liability for Harm vs Ex Ante Safety Regulation: Substitutes guard Complements? The American Economic Review, 80(1), 888-901.
Kornai, János (1992): The Socialist System: The Political Economy of Communism. Princeton, NJ: Princeton University Press.
Mefford, Aron (1997): Lex Informatica: Foundations of Law on the Internet. Indiana Journal of Global Legal Studies, 5(1), 211-238.
Mises, Ludwig von (1944): Bureaucracy. New Haven: Yale University Press.
O'Riordan, Tim - Cameron, James eds. (2002): Interpreting the Precautionary Principle. London: Routledge.
Phillipson, Gavin (1999): The Human Rights Act, Horizontal Effect and the Common Law: a Bang or a Whimper. Modern Law Review, 62(6), 824-849. Online: https://doi.org/10.1111/1468-2230.00240
Polanyi, Karl (2001): The Great Transformation. The Political and Economic Origins of Our Times. Boston, MA: Beacon Press.
Raz, Joseph (1990): Practical Reason and Norms. Princeton, NJ: Princeton University Press.
Rehbinder, Manfred (1986): Die Begründung der Rechtssoziologie durch Eugen Ehrlich. Berlin: Dunckert & Humblot.
Reidenberg, Joel R. (1997): Lex Informatica: The Formulation of Information Policy Rules through Technology. Texas Law Review, 76(3), 553-594.
Schlesinger, Philip (2022): The Neo-Regulation of Internet Platforms in the United Kingdom. Policy & Internet, 14 (1), 47-62. Online: https://doi.org/10.1002/poi3.288
Senden, Linda (2005): Soft Law, Self- regulation and Co-regulation in European Law: Where do they Meet? Electronic Journal of Comparative Law, 9(1). Online: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=943063
Stark, David - Pais, Ivana (2020): Algorithmic Management in the Platform Economy. Sociologica, 14(3), 47-72. Online: https://doi.org/10.6092/issn.1971-8853/12221
Sunstein, Cass (2005): Laws of Fear. Beyond the Precautionary Principle. Cambridge: Cambridge University Press. Online: https://doi.org/10.1017/CBO9780511790850
Susskind, Richard (2009): The End of Lawyers? Rethinking the Nature of Legal Services. Oxford: Oxford University Press.
- 123/124 -
Tushnet, Mark (2003): The Issue of State Action/Horizontal Effect in Comparative Constitutional Law. International Journal of Constitutional Law, 1(1), 79-98. Online: https://doi.org/10.1093/icon/1.1.79
Yeung, Karen - Lodge, Martin (2019): Introduction. In Yeong, Karen - Lodge, Martin (eds.): Algorithmic Regulation. Oxford: Oxford University Press. Online: https://doi.org/10.1093/oso/9780198838494.003.0001
Van Dijck, José - Nieborg, David - Poell, Thomas (2019): Reframing Platform Power. Internet Policy Review, 8(2). Online: https://doi.org/10.14763/2019.2.1414
Vicarelli, Fausto (1984): Keynes: The Instability of Capitalism. [s. l.]: University of Pennsylvania Press. Online: https://doi.org/10.9783/9781512819892
Walker, Neil (2005): Legal Theory and the European Union European University Institute. EUI Working Papers, Law 2005/16. Online: https://doi.org/10.2139/ssrn.891032
Ződi, Zsolt (2022): Algorithmic Explainability and Legal Reasoning. The Theory and Practice of Legislation, 10(1), 67-92. Online: https://doi.org/10.1080/20508840.2022.2033945
Commission Delegated Regulation (EU) 2019/945 of 12 March 2019 on unmanned aircraft systems and third country operators of unmanned aircraft systems
Communication Decency Act of the USA
Council Directive 93/13/EEC (April 5, 1993) on unfair terms in consumer contracts
Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market ('Directive on electronic commerce')
Directive 2010/13/EU of the European Parliament and of the Council of 10 March 2010 on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the provision of audiovisual media services (Audiovisual Media Services Directive - old AVMSD)
Directive (EU) 2018/1808 of the European Parliament and of the Council of 14 November 2018 amending
Directive 2010/13/EU on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the provision of audiovisual media services (Audiovisual Media Services Directive) in view of changing market realities
Directive (EU) 2019/790 of the European Parliament and of the Council of 17 April 2019 on copyright and related rights in the Digital Single Market and amending Directives 96/9/EC and 2001/29/EC (Copyright Directive)
Network-enforcement law (Netzwerkdurchsetzungsgezetz - NetzDG) of Germany
- 124/125 -
Proposal for a Directive of the European Parliament and of the Council on improving working conditions in platform work COM(2021) 762 final (Platform work proposal). Online: https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=COM%3A2021%3A762%3AFIN
Proposal for a regulation of the European Parliament and of the Council laying down harmonised rules on artificial intelligence (Artificial Intelligence Act) and amending certain union legislative acts
Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation - GDPR)
Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20 June 2019 on promoting fairness and transparency for business users of online intermediation services (Platform-to-Business Regulation - P2B)
Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (Digital Markets Act - DMA)
Regulation (EU) 2022/2065 of the European Parliament and of the Council of 19 October 2022 on a Single Market For Digital Services and amending Directive 2000/31/EC (Digital Services Act - DSA)
Shelley v. Kraemer, 334 U.S. 1 (1948)
Zsolt Ződi is a Senior Researcher at the Ludovika University of Public Service. He graduated as a lawyer from the University of Miskolc in 1991, where he then taught legal theory until 1995. Then he worked in the legal publishing business for 20 years as an editor, publisher and publishing director in various multinational and domestic legal content providers. In 2012, he was awarded a PhD by the University of Pécs with a dissertation on legal informatics. Since then, he has worked and taught at the Legal Studies Institute of the HUN-REN Hungarian Research Network, Eötvös Loránd University, University of Miskolc, and in Corvinus University of Budapest both legal theory and info-communication law. He habilitated at the Faculty of Law in Eötvös Loránd University in 2021. He is the author of 3 books and more than 110 articles. ■
NOTES
[1] See e.g. the main regulatory tool of the EU between 2000 and 2022, Directive 2000/31/EC ('Directive on electronic commerce').
[2] A different approach was employed in 2017 with the German "Network-enforcement law" (Netzwerkdurchsetzungsgesetz - NetzDG), which imposed certain transparency and procedural obligations on the social media platforms.
[3] In the second half of the decade, similar initiatives appeared in other parts of the world. Thus, in the USA, several federal and state-level legislative drafts have been published aimed at regulating the platform scene. These are very similar to their European counterparts in terms of their goals, and sometimes also in their details, so the theoretical considerations described here also apply to them.
[4] Directive 2010/13/EU (Audiovisual Media Services Directive - old AVMSD) and its revision, Directive (EU) 2018/1808 (Audiovisual Media Services Directive).
[5] Directive (EU) 2019/790 (Copyright Directive).
[6] Regulation (EU) 2019/1150 (Platform-to-Business Regulation - P2B).
[7] Regulation (EU) 2022/2065 (Digital Services Act - DSA).
[8] Regulation (EU) 2022/1925 (Digital Markets Act - DMA).
[9] Information page: https://ec.europa.eu/commission/presscorner/detail/en/ip_21_6605; Text of draft: Proposal for a Directive of the European Parliament and of the Council on improving working conditions in platform work COM(2021) 762 final (Platform work proposal).
[10] These are the Austrian legal sociologist, Eugen Ehrlich's expressions. He was the first to point out that in addition to the "decision law" (Entscheidungsnormen) applied in (and by) the courts, there is another huge body of law, the "living law", which is used by economic participants and private individuals in their everyday (economic) life. This living law has something to do with ex ante regulation, but it is not identical to it. At the same time, ex ante law can also be well contrasted with the ex-post (lawyers') law. See Ehrlich 2022: 210, 240; Rehbinder 1986: 50.
[11] Kornai 1992.
[12] Kornai 1992: 91-109. Kornai's main source of inspiration was Karl Polanyi, (Polanyi 2001: 45) who called these mechanisms "integration schemes". Julie Cohen (Cohen 2017: 15) and Stark and Pais (Stark-Pais 2020) depart also from Polanyi's theory when explaining platforms. Both authors come to conclusions that are very similar to the ones here. In the same vein: Mises 1944.
[13] According to Keynes (Keynes 1936: 313-332), the inherent property of the market is that it operates in trade cycles and occasionally drifts into crises. See also Vicarelli 1984.
[14] This is the core of Montesquieu's theory.
[15] Some examples from the rich literature: Gardbaum 2003; Phillipson 1999; Tushnet 2003.
[16] Shelley v. Kraemer, 334 U.S. 1 (1948).
[17] Bibó 1986: 120.
[18] Hildebrandt 2018.
[19] The opening sentence of the official website of the European Commission introducing the DSA package is: "The Digital Services Act and Digital Markets Act aim to create a safer digital space where the fundamental rights of users are protected and to establish a level playing field for businesses." https://digital-strategy.ec.europa.eu/en/policies/digital-services-act-package Schlesinger characterises the British situation in the following way: "The British policy agenda and regulatory response are presently encapsulated in two portmanteau rallying points: remedying 'online harms' (which mainly encompass social and political issues) and pursuing a 'procompetition' approach (by addressing malfunctioning markets to promote consumer interests and business innovation)" Schlesinger 2022.
[20] Recently, the main legal source of the platform regulation is the Communication Decency Act, Section 230, and approximately one dozen federal and state bills are on the table of legislators across the U.S.
[21] Sunstein 2005: 14.
[22] H.R.3849 - 117th Congress (2021-2022), see: www.congress.gov/bill/117th-congress/house-bill/3849/text
[23] Regulation (EU) 2016/679, (General Data Protection Regulation-GDPR) Article 20.
[24] Walker 2005.
[25] Fried 2003.
[26] An overview of the history of the precautionary principle: Harremoës et al. 2001 and O'Riordan 2002.
[27] Sunstein 2005: 13. Some say the principle has a different origin: according to Christiansen, the principle originates from the German Vorsorgenprinzip of the 1930s. Boehmer 2002.
[28] Kolstadt et al. 1990.
[29] Directive 2000/31/EC of the European Parliament and of the Council, Article 14: "Hosting/ 1. Where an information society service is provided that consists of the storage of information provided by a recipient of the service, Member States shall ensure that the service provider is not liable for the information stored at the request of a recipient of the service, on condition that: (a) the provider does not have actual knowledge of illegal activity or information and, as regards claims for damages, is not aware of facts or circumstances from which the illegal activity or information is apparent; or (b) the provider, upon obtaining such knowledge or awareness, acts expeditiously to remove or to disable access to the information."
[30] DSA Article 16.
[31] See Darley et al. 2001.
[32] Raz 1990: 16.
[33] Cass Sunstein argued convincingly against excessive ex ante (precautionary) regulation. Sunstein 2005.
[34] See Kolstadt et al. 1990.
[35] Galle 2015.
[36] For example, Article 9 of the draft DSA, which regulates the transfer of information to the authorities, in its final version is almost double the length of the original text of the proposal, partly due to the increasingly detailed obligations placed on the authorities, and partly of the intermediary service providers. The same is also true for Articles 12, 17 and many other texts.
[37] Senden 2005.
[38] European Commission 2001.
[39] European Commission 2002.
[40] This expression is used by Caroline Cauffman and Catalina Goanta in connection with the DSA. Cauffman-Goanta 2021.
[41] Brownsword 2011 and Brownsword 2005.
[42] Brownsword 2021.
[43] Brownsword 2021: 23.
[44] Brownsword 2021: 32
[45] Originally by Joel R. Reidenberg (Reidenberg 1997), which was later adopted by others (e.g. Mefford 1997), but the idea - the increasing role of codes in influencing behaviour in cyberspace - really became widely known with Lessig's theory. Interestingly, Karen Yeung and Martin Lodge are giving an entirely different origin story for the regulation by algorithms in their writing (Yeung-Lodge 2019: 4), stating that algorithmic regulation "was popularized by Silicon Valley entrepreneur Tim O'Reilly in 2013" but "the idea that computational algorithms might be understood as a form of social ordering was proposed some time earlier by sociologist A. Aneesh".
[46] Susskind 2009: 141.
[47] Brownsword et al. 2016.
[48] DSA Article 20.
[49] DSA Article 27(3) (adjustability of recommendation systems) Article 38 (selectability of recommendation systems not based on profiling).
[50] E.g. DSA Article 9(2) a) "redress mechanisms", Article 40(8) d) "technical and organizational measures".
[51] DSA Article 14.
[52] DSA Article 44.
[53] Ződi 2022.
[54] Joh 2016.
[55] Artificial Intelligence Act.
[56] There are two drafts, the Data Governance Regulation and the Data Sharing Regulation: European Commission 2020; European Commission 2022.
[57] Van Dijck et al. 2019.
[58] Jürgen Habermas's expression. Habermas 1987 [1981].
[59] "[P]urchasers of goods and users of services must be protected against abuse of power by the seller or the service provider." Council Directive 93/13/EEC - Consumer Protection Directive, Recital 9.
[60] See Commission Delegated Regulation (EU) 2019/945 Annex, part 3, paragraph (15), point a).
[61] Council Directive 93/13/EEC.
[62] DSA Article 14.
[63] Dumbrava 2021.
[64] "The DSA and DMA have two main goals: [...] to create a safer digital space, [...] and to establish a level playing field for businesses." See: https://digital-strategy.ec.europa.eu/en/policies/digital-services-act-package
[65] See for example Facebook's policy on dangerous individuals and organisations: https://transparency.fb.com/hu-hu/policies/community-standards/dangerous-individuals-organizations/
[66] DSA Article 20.
[67] DSA Article 16.
[68] DSA Article 14, P2B Article 3.
[69] DSA Article 14(1).
[70] DSA Article 23.
[71] DSA Article 27.
[72] P2B Article 5.
[73] P2B Article 3(1) c).
[74] DSA Article 17, P2B Article 4.
[75] DSA Article 20.
[76] P2B Article 3.
[77] P2B Article 4.
[78] Platform work draft Article 8 of the draft directive.
[79] DSA Article 27.
[80] DSA Article 27.
[81] P2B Article 5.
[82] Platform work draft Articles 6-10.
[83] Platform work draft Articles 6-10.
[84] P2B Article 12.
[85] DSA Article 20.
[86] DSA Article 15 Transparency reporting for the intermediary services, Article 24 for platforms, Article 39 for advertising transparency on VLOPs, and Article 42 for transparency reporting of VLOPs.
Lábjegyzetek:
[1] The author is Senior Researcher, Ludovika University of Public Service, Institute of the Information Society, zodi.zsolt@uni-nke.hu
Visszaugrás
