Document Actions

Articles

Fixing Copyright Reform: A Better Solution to Online Infringement

  1. Dr. Christina Angelopoulos
  2. Dr. João Pedro Quintais

Abstract

The newly-adopted Directive on Copyright in the Digital Single Market (DSMD) will fundamentally reshape EU copyright law. Among its most controversial offerings is Article 17, the so-called “value gap” provision, aimed at solving the alleged mismatch between the value that online content-sharing platforms extract from creative content and the revenue returned to the copyright-holders. This article argues that the new rules are misguided, misconceiving the real problems afflicting modern copyright. These are the proliferation of copyright infringement online in general – not only through content-sharing platforms – and the current piecemeal harmonisation of the rules on the liability of the intermediaries whose services are used to access and disseminate copyright-protected content. The current outdated and fragmented EU legal framework is ill-equipped to address these problems. Instead, it creates legal uncertainty for users and intermediaries in the online environment, while also failing to compensate creators fairly. The new rules will not change this. This article examines the pre-DSMD acquis and proposes a better solution than Article 17, consisting of two key changes: (a) the introduction of a harmonised EU framework for accessory liability for third party copyright infringement; and (b) the adoption of an alternative compensation system for right-holders covering non-commercial direct copyright use by the end-users of certain online platforms.

Keywords

1. Introduction*

1

In September 2016, the European Commission published its proposal for a new Directive on Copyright in the Digital Single Market (DSMD). [1] The proposal was controversial from the start. Almost every step of the legislative process was the subject of intense lobbying and debate, up until the final text was approved in 2019. [2]

2

Among the more controversial offerings of the Directive is Article 17 (13 in earlier drafts) on the “[u]se of protected content by online content-sharing service providers”. Although neither the Explanatory Memorandum nor the Impact Assessment to the Directive explicitly use the term, it is widely accepted that the main driver behind Article 17 is what has become known as the “value gap”, i.e. the alleged mismatch between the value that online sharing platforms extract from creative content and the revenue returned to the copyright-holders.

3

Article 17 regulates “online content-sharing service providers” (OCSSPs). These are defined as service providers with a profit-making purpose that store and give the public access to large numbers of copyright works uploaded by their users, which they organise and promote. This applies not only to well-known platforms, like YouTube, Facebook or Vimeo, but also to any type of user-upload platform that fits this broad definition and is not expressly excluded. [3] The provision changes existing law by stating that such platforms carry out acts of communication to the public when they give access to works uploaded by their users. [4] These platforms are therefore made primarily liable for their users’ uploads. Under the new Directive, OCSSPs no longer benefit in relation to such uploads from the hosting safe harbour, which was previously available to many of them. [5]

4

Platforms that fall under this regime have two options. First, they may obtain authorisation from the right-holders by concluding a licensing agreement for the content uploaded by users. [6] This is problematic, as it’s nearly impossible to envision a scenario where a platform of this type can obtain all the required licences for the potentially millions of works uploaded by their users. [7] The second option provided is therefore likely to be heavily relied upon. According to this, platforms can avoid liability if they meet a number of cumulative conditions/requirements. They must demonstrate that they have: (a) made best efforts to obtain an authorisation; (b) made best efforts to ensure the unavailability of specific works for which the right-holders have provided them with the relevant and necessary information; and (c) acted expeditiously, subsequent to a notice by the right-holders, to take down or disable access to infringing content and made best efforts to prevent its future upload. [8]

5

Many commentators argued during the legislative process that the obligations introduced by Article 17 would be incompatible with pre-existing EU directives, as well as with the EU Charter of Fundamental Rights, [9] as interpreted by the CJEU. [10] In order to ensure the unavailability of works and escape liability, platforms will be obliged to examine all uploaded content to ensure that it does not infringe copyright in works listed by right-holders and remove that which does. Given the immense amounts of content that many platforms store, in practice this will involve the use of automatic content recognition technologies, otherwise known as “upload filters”. Critics point out that, whether done manually or through automated tools, such checks will require general monitoring, an outcome incompatible with earlier law [11] and which the directive explicitly rejects. [12] Crucially, in its case law, the CJEU has also stated that broad filtering measures would be in violation of platforms’ freedom to conduct a business, as well as users’ rights to the protection of personal data and freedom of expression. [13]

6

To avoid the negative effects of such obligations, the provision requires platforms to ensure filters will not prevent the availability of works which do not infringe copyright. This would include uses covered by an exception or limitation, such as those for the protection of quotation, criticism, review, caricature, parody, or pastiche. [14] Unfortunately, existing content recognition technologies are incapable of accommodating such dynamic and context-specific exceptions. Human review is unlikely to help much, unless undertaken by judges. The result will be that many otherwise lawful uses will be blocked. [15] Although the DSMD does require that platforms implement complaint and redress mechanisms for users, these are unlikely to mitigate these effects in any meaningful way, since these mechanisms are typically ineffective. [16] Article 17 also incorporates some safeguards for providers’ freedom to conduct a business, in the form of factors to be taken into account in determining whether an OCSSP has complied with its obligations. [17] Whether these are sufficient will be a question for the courts. Finally, while the Directive does explicitly state that it should not lead to the unlawful processing of personal data, [18] it is not obvious how the filtering that Article 17 incentivises can avoid that outcome. [19]

7

In addition to these shortcomings, this article suggests that Article 17 suffers from a more fundamental weakness: it is based on a misconception of the real problem afflicting EU copyright law, i.e. the proliferation of copyright infringement online in general, not only through Web 2.0 hosts. This problem is compounded by an increasingly outdated EU copyright framework that allows infringing end-users to hide behind their online anonymity, while failing to provide any mechanism for the compensation or remuneration of right-holders for the infringements these users commit. [20] Faced with this impasse, right-holders have shifted their focus to internet intermediaries. Yet, while the CJEU’s recent case law has waded into the tricky area of intermediary liability, no complete system of rules determining what obligations intermediaries have to prevent or remove online copyright infringement currently exists at the EU level.

8

Absent a more stable legal basis, targeted superstructure initiatives such as the new directive are set up for failure. At best they can improve a small part of a larger problem and at worst introduce significant legal uncertainty. If EU copyright law is to be reformed, it is on the above crucial weak spots that proposals should focus. This article proposes an alternative approach to Article 17, intended to better target these weak spots. Based on our pre-existing research into the areas of intermediary liability and copyright compensation systems, we advance a two-pronged proposal, consisting of: (a) the introduction of a harmonised EU framework for accessory liability for third party copyright infringement; and (b) the adoption of an alternative compensation system for right-holders covering non-commercial direct copyright use by the end-users of certain online platforms. [21] As we explain, this solution avoids the difficulties encountered by Article 17 DSMD, while successfully targeting the copyright framework’s real failings.

9

The article progresses in the following manner. Part 2 begins with a descriptive account of the relevant pre-DSMD EU law on copyright, in particular the Information Society Directive (InfoSoc Directive) [22] and the E-Commerce Directive (ECD), [23] as interpreted by the CJEU. Part 3 moves on to the article’s normative offering: our proposed alternative approach to copyright reform. Here, each of the proposal’s two components are analysed in detail. Part 4 concludes with a brief pragmatic assessment.

2. Identifying the Real Problems: Two Crucial Weaknesses of the Current Copyright Acquis

10

Before a solution to the problem of online copyright infringement can be devised, a proper understanding of the current legal framework governing the issue is necessary. Such an understanding can enable the identification of that framework’s failings: if online copyright infringement is a problem, in what ways is the current law struggling to address this? Once the law’s weak points have been located, targeted solutions can be deployed. This section examines the pre-DSMD EU copyright framework from two perspectives: firstly, the legal position of individual infringers and, secondly, the legal position of the intermediaries whose services such infringers use. In this way, the inadequacies of the current rules’ response to the two main contributors to modern online copyright infringement can be identified.

2.1. The Precarious Position of Individual Users

11

The first crucial weakness of the current acquis is found in the precarious position in which it places individual users. The everyday practices of such users on the Internet involve a diverse array of acts, including browsing, downloading, streaming, stream ripping, uploading, and hyperlinking. These activities may involve making copies of and disseminating copyright-protected content online and, as a result, trigger the application of the exclusive rights of reproduction and communication to the public, as harmonised by Articles 2 and 3 of the InfoSoc Directive.

12

In particular, following CJEU decisions, the right of reproduction applies to the online consumption, manipulation and dissemination of content by end-users, as these activities require the making of digital copies. [24] This broad interpretation may be questioned, with commentators suggesting that it is inappropriate for the digital age. [25] Concerning the right of communication to the public, the Court has developed a complex set of criteria. [26] In simple terms, a use will amount to a communication to the public if it meets two cumulative conditions. It must constitute: (a) an “act of communication”, that is made to (b) “the public”. To this effect, the Court carries out an “individualised assessment” of several “autonomous and interdependent” criteria in light of the circumstances of each case. [27] Among others, the Court looks at the “role of the user”, his “deliberate/intentional intervention” and whether the communication at stake is through a “separate”/“specific” technical means or directed at a “new public”. [28] Cumulative audiences amount to a “public”, as long as they are not “insignificant”. Finally, whether or not the “communication” is of a profit-making nature also has to be considered. [29] These criteria do not always neatly apply to the online use of content, a problem compounded by the fact that the legal status of certain acts may depend on the imposition of contractual and technical restrictions by right-holders. [30] For example, it is not currently clear whether sharing a work with a restricted group of “friends” on a social network would amount to a communication to the public. Moreover, as the CJEU expands and contracts the scope of the right from case to case, greater flexibility might be injected into judicial deliberations, but end-users themselves are placed on legal quicksand.

13

Ultimately, as a result of the EU legal framework on reproduction and communication to the public, a significant number of individuals’ online activities either prompt the application of exclusive rights or are legally ambiguous. In principle, the reproduction right applies to browsing, downloading, the reception of streams from unauthorised sources, stream ripping, certain uploads, and the making of digital adaptations. The right of communication to the public is relevant for the online transmission of streams, uploading to publicly accessible websites, and many types of hyperlinking. [31]

14

Of course, even where exclusive rights apply, a use may nevertheless be allowed by the exceptions set out in Article 5 of the InfoSoc Directive. This, however, is not generous to users, as it excludes rights in software and databases, defines most exceptions as optional for Member States, [32] and allows limitations to be overridden by contract, as well as the use of technological protection measures. [33] The EU case law on exceptions mostly follows the canon of strict interpretation, meaning that their scope is narrowly defined. [34] The result is that existing exceptions provide little coverage for the online practices of users.

15

Finally, little relief can be found in the legal regulation of digital adaptations. At the EU level, there is no harmonised right of adaptation. National law may contain a separate right of adaptation, as well as exceptions providing breathing space for the making and dissemination of derivative works. [35] But such space can only be limited for two reasons. First, the broad interpretation of the harmonised exclusive right of reproduction by the CJEU causes it to extend to most literal and non-literal reproductions that form part of the derivative work. Second, where that is the case, the narrow interpretation of the harmonised exceptions (quotation, incidental inclusion, and parody) excludes their application to many digital adaptations. [36] Thus, to create such a space, legislative reform at EU level is needed, e.g. through the creation of a specific exception for user-generated content. [37]

16

As a result of the above, the current framework leaves individual users in a precarious legal position, while seemingly also conflicting with the objectives of EU copyright law. [38] It does not promote legal certainty, the smooth functioning of the internal market, the development of new technologies, or foster the circulation of culture. It erodes the legitimacy of copyright law by divorcing it from social norms. Finally, assuming that a significant number of online infringments by individuals is unenforceable – either de facto or due to limits imposed by fundamental rights [39] – the expansion of exclusivity in the online environment fails to ensure an appropriate remuneration to right-holders for the (unauthorised) use of works and, consequently, an adequate level of protection: [40] right-holders may be protected, but only in theory.

2.2. The Partial Harmonisation of Intermediary Liability

17

Most online copyright infringements are not performed by persons relying exclusively on their own technical capabilities. Instead, they use the services of internet intermediaries. In addition to that of individual users therefore, a second significant question concerns the liability of the providers of such services. When should they be held liable for copyright infringements committed by end-users employing their services? At the moment, no complete EU system on intermediary liability in copyright exists.

18

Prior to the adoption of the new directive, the most relevant EU provisions were the so-called “safe harbours”. These are conditional immunities introduced in the E-Commerce Directive (ECD) that shield intermediaries from liability under specific circumstances. While the safe harbours do not harmonise intermediary liability per se, they do set limits to its application. For the purposes of this article, the hosting safe harbour of Article 14 ECD is especially relevant. This provides protection to intermediaries that store information posted by third parties, as long as the host does not have actual or constructive knowledge that the information is illegal or, upon obtaining such knowledge, removes it. [41] Many of Article 17’s OCSSPs had been presumed to be covered by the provision.

19

Complementing the safe harbours, Article 15(1) ECD prohibits Member States from imposing general obligations on protected providers to monitor the information that they transmit or store, or to seek facts or circumstances indicating illegal activity. In combination with Article 14, this results in the effective introduction into EU law of what is known as a “notice-and-takedown” (NTD) regime, i.e. a system that obliges hosting providers to remove infringements after receiving notification of their illegality or lose the safe harbour.

20

Importantly, even if a platform enjoys safe harbour protection, this will not necessarily mean that no action can be taken against it. Article 8(3) of the InfoSoc Directive instructs Member States to ensure that injunctions remain available against intermediaries whose services are used by third parties to infringe copyright or related rights. The general monitoring prohibition of Article 15(1) ECD sets limits here as well. The provision has been interpreted by the CJEU as excluding injunctions obliging intermediaries to monitor actively all the data they transmit or store. [42]

21

The CJEU has also approached the issue of injunctions against intermediaries through the lens of fundamental rights. The Court has held that such injunctions involve a clash between opposing fundamental rights. According to consistent case law, such clashes should be resolved by the Member States by striking a “fair balance” between all relevant fundamental rights. [43] As a general rule, in cases of copyright and related rights enforcement, a tripartite schematic focus on copyright holders, intermediaries and internet users can be identified. For the first, the right to intellectual property (Article 17(2) Charter) is at stake. Depending on the circumstances, in the application of injunctions this may collide with the intermediaries’ freedom to conduct a business (Article 16 Charter). In addition, the rights of end-users to the protection of their personal data (Article 8 Charter), their private life (Article 7 Charter) and their freedom to receive and impart information (Article 11 Charter) must be taken into account. The Court has found that filtering obligations would not strike a ‘fair balance’ between all relevant fundamental rights. [44] On the other hand, more moderate requests are more likely to be granted. [45]

22

In addition to the above, the only harmonised source of intermediary liability rules in EU law can be found in the regular provisions on primary copyright infringement. Naturally, these may be applied to hold Internet intermediaries liable for the infringements they perform in the same way as any other person. Yet, in recent case law, the CJEU has reshaped them to do something more: hold defendants liable for infringements performed by others. In this way, their bearing on intermediary liability has significantly increased.

23

The relevant jurisprudence has so far focused on the right of communication to the public of Article 3 of the InfoSoc Directive. The CJEU’s aforementioned “individualised” approach to this right has allowed it to gradually relax its early finding that an “act of communication” requires an “indispensable” intervention on the part of the defendant, such that the relevant public would not otherwise be able to enjoy the work. [46] Instead, the Court now accepts that it is sufficient that the public would be able to do so only with difficulty. [47] The consequences of this expansion are controlled through the requirement that the relevant public be a “new” one. In the recent hyperlinking decision of GS Media, the notion of “new public” was connected with the defendant’s knowledge of the infringement or profit derived from it. [48]

24

This evolution has proved significant for intermediary liability. In June 2017, 10 months after the publication of the Commission’s proposal for the new directive, GS Media’s knowledge-based logic was applied in Ziggo to find that the peer-to-peer file-sharing website The Pirate Bay reaches a “new public” and thus communicates works to the public. [49] In this way, it was confirmed that the mental element GS Media introduced is not intended to apply exclusively to hyperlinking, but has a wider applicability. The results could be particularly important for the host service providers targeted by Article 17. After all, The Pirate Bay is in effect a host, albeit of torrent files and magnet links, as opposed to content files – and if links can constitute infringements, the difference between the two is minimal. Moreover, knowingly hosting a link to infringing content is arguably a more tenuous connection than knowingly hosting the infringing content itself. As commentators have observed, in this way, the CJEU appears to have pre-empted the EU legislator and introduced its own positively-stated rules on intermediary liability. [50]

25

Notably, the CJEU’s approach rests on a curious blending of the notions of primary and accessory liability in copyright. Primary infringement occurs where a person engages in a copyright infringement themselves. Accessory infringement occurs where one person participates in – i.e. acts as an accessory to – a copyright infringement performed by another. While knowledge is traditionally a hallmark of accessory liability in copyright law, [51] primary infringement has always been understood to be a strict liability tort. [52]

26

Obfuscating the division between the two areas can be tricky. Most obviously, the consistent requirement of a mental element would mean turning primary copyright infringement into a fault-based tort. It is however doubtful that this was the CJEU’s intention. What is far more likely is that the CJEU will, in future, rely on the variable normative power it has assigned to the multiple interdependent criteria on which it establishes communication to the public to conclude that the condition of knowledge is only necessary where the infringement depends on the existence of a prior infringement by another. That is to say, knowledge will matter only in cases where the defendant was acting as an accessory, as opposed to the primary perpetrator. [53]

27

Questions however remain: what type of involvement in the infringement of another is sufficient? What is the appropriate mental element? Whether the exceptions and limitations of Article 5 of the InfoSoc Directive will apply to parties whose infringement depends on their knowledge of the infringement is also unclear. More importantly, although it represents a definite step in the right direction, the CJEU’s approach stops short of the full harmonisation of accessory liability in copyright. If an intermediary cannot reasonably be found to be communicating to the public, assuming no safe harbour applies, a pan-EU solution will remain elusive: national rules on accessory liability might result in different outcomes for the same defendant in different Member States.

28

Article 17 does not address these problems. For one thing, it only applies to OCSSPs. Furthermore, its relationship to the CJEU case law on communication to the public is unclear. The new provision states that OCSSPs communicate the uploads of their users to the public. But what does this mean for the CJEU’s careful interpretation of this right? Do the Court’s “autonomous and interdependent” criteria remain relevant or have they now been made redundant, whether for OCSSPs or generally? The new rules create more questions than they answer.  [54]

3. A Better Way Forward: A Two-pronged Approach to Online Infringement

29

As the above analysis reveals, the current EU law on copyright infringement has major weaknesses. It provides little breathing space for individuals to use protected content with legal certainty, emphasises exclusivity and enforcement over remuneration and has evolved into a complex legal maze regarding the liability of online platforms. The EU legislator’s “value gap”-driven and OCSSP-focused solution will not solve these problems. It is likely incompatible with existing directives and with the Charter, increases legal uncertainty, and fails to take the interests of creators into proper account. [55]

30

This article argues that copyright reform should instead focus on the weaknesses identified above in the current EU law on copyright infringement. It recommends two mechanisms for overcoming these weaknesses and achieving real improvement of EU copyright law for all involved stakeholders. These would be complementary and intended to support each other through their parallel implementation. The first is the adoption of a truly harmonised EU framework for accessory liability for copyright infringement. The second is the introduction of an alternative compensation system for right-holders covering non-commercial direct copyright infringement by end-users online, in particular in the context of user-upload platforms.

31

Below, we examine how this two-pronged approach would work and what each of these mechanisms would contribute to the ultimate solution to the problem of online copyright infringement. Subsequently, the anticipated effect of their combined adoption on the operation of EU copyright law is briefly discussed.

3.1. Harmonisation of accessory copyright liability

32

As Section 2.2. suggests, one of the most prominent gaps in the edifice of modern European copyright law is the lack of a harmonised regime for accessory liability. [56] In view of the ubiquity of intermediation in internet-based communications, this omission has become increasingly glaring. The rules of accessory liability are those suitable to govern questions of intermediary liability, i.e. when an intermediary should be held liable for the infringements of third parties (most commonly, its users). The lack of a harmonised regime has therefore resulted in the fragmentation of the legal framework in an area that is particularly relevant to online infringement. Introducing a harmonised solution would overcome this fragmentation and create a sound legal basis for addressing such infringement. Once adopted, this solution could also be applied to the online sharing platforms implicated in the “value gap” controversy, thus providing a more theoretically consistent solution to their liability than that offered by Article 17.

33

As an added advantage, such harmonisation would also help achieve greater clarity with regard to the European rules of primary liability in copyright, by decisively alleviating the pressures for their over-expansive judicial interpretation. Ideally, the harmonised framework would properly distinguish accessory from primary liability. While the current “blended” approach taken by the CJEU has the advantage of enabling harmonisation through case law, legislative intervention would make such expediencies redundant. At the same time, lessons can be learnt from the Court’s recent decisions in the area. The legislator is certainly free to modify existing EU secondary law as it sees fit, as long as it respects the Charter. Yet undercutting established case law undermines the development of a coherent system. Instead of grafting supervenient provisions like Article 17 on to EU copyright law, the EU legislator would be wise take its cue from existing case law.

34

Examining this case law reveals the outline of a convincing framework for a European accessory copyright liability system. As seen above, while traditionally copyright infringement has relied only on the examination of the defendant’s conduct, in its decisions on accessory liability, the CJEU has added a new mental element. Happily, this approach accords with the usual schematic relied upon in countries with existing accessory liability doctrines. As a rule, in such regimes accessory liability depends on two essential elements: a “conduct element” and a “mental element”. [57] We suggest that a future EU intermediary liability copyright regime should follow a similar approach. While the first element should focus on whether the defendant’s behaviour has contributed to a copyright infringement, the second should concern their inner involvement in the infringement, i.e. whether they demonstrated the mindset of a culpable party. Where at least one of these elements is absent, the defendant should be absolved from any liability. If both are satisfied, two possibilities arise: either the defendant should immediately be held liable for the infringement; or they are placed under an obligation to take appropriate action.

35

Careful consideration of the right standard for both elements is necessary, as well as of the appropriate consequences where they are both met. Again, the CJEU’s case law can guide the analysis. Where ambiguity remains, other sources of legislative inspiration should be considered. In this regard, one important question concerns the relevance of the ECD’s safe harbours. It may seem logical that a well-crafted defence should do more than mirror the provisions that determine infringement. We suggest however that the safe harbours should not be approached in this way. They were adopted not to provide exceptions to the exclusive rights of copyright holders, but to introduce harmonised horizontal rules for the smooth functioning of the single market. [58] It is therefore appropriate to treat them as guidelines for a future harmonised EU accessory liability regime for copyright. [59] By the same token, Article 15 ECD, as well as the Court’s existing case law on “fair balance” and injunctions should also be taken into consideration. The proposal below draws heavily from all three of these sources.

3.1.1. The Conduct Element of Accessory Liability

36

As explained in Section 2.1., the CJEU’s approach to communication to the public has gradually evolved over the past decade. Helpfully, recent case law has planted the seeds for a harmonised system for accessory liability in copyright. It can therefore be relied upon to guide the construction of the elements of a harmonised European accessory liability regime. A more detailed examination is thus appropriate. In this section, we start with the conduct element. The Court’s decisions on the notion of an “act of communication” are helpful in tracing this out. As noted above, the “individualised” approach adopted by the CJEU has allowed it gradually to relax its approach.

37

Notably, the majority of the CJEU’s rulings in this area deal with what might be termed “secondary use”. They thus concern either retransmissions of broadcasts by wire or wireless means (see e.g. ITV Broadcasting, SBS Belgium, and zürs.net) [60] or the playing of phonograms or broadcasts in a public space (see e.g. SGAE, Divani Akropolis, Del Corso, Phonographic Performance (Ireland) and Reha Training). [61] In these cases, the Court’s analysis still remains closely tied to the classic language of a “transmission”, that is embedded in Recital 23 of the InfoSoc Directive. [62] According to this, the right of communication to the public should not cover any acts other than the transmission or retransmission of a work to the public by wire or wireless means. However, a wrinkle then appears. Starting with Svensson, cases come before the CJEU that concern acts of secondary use rather than secondary infringement, i.e. acts whose infringing nature depends on the existence of a primary infringement by another. The Court accepted these as valid acts of communication. GS Media, Filmspeler  [63] and Ziggo continued this trend, fleshing out the relevant principles.

38

The expansive tendency of the Court was enabled by the language of a “deliberate/intentional intervention to give access”, already adopted in previous decisions. The notion of an “intervention”, allowed the Court to move discretely beyond “transmissions”, broadening the scope of the restricted act. So, for example, Svensson incorporates within the notion of an act of communication “an intervention without which users would not be able to access the works transmitted” by another, thus apparently differentiating between “transmissions” and other “acts of communication”. In Ziggo, the Court confirms this approach by stating that an act of communication “covers” (but apparently is not limited to) any transmission or retransmission of a work. [64] Similarly, it is “sufficient” that the work is made available to the public. [65] The notion of an “act of communication” is ultimately defined as “any act by which a user, with full knowledge of the relevant facts, provides its clients with access to protected works”.  [66] These decisions reveal the appropriate scope of the conduct element in a harmonised EU accessory liability regime: it seems reasonable to conclude that the notion of an “intervention” that does not amount to a “transmission” [67] is where a harmonised accessory liability regime would come into operation.

39

Of course, not any “intervention” will suffice. As noted above, the Court had initially required that the intervention must be “indispensable” [68] or at least “essential” [69] for the dissemination of the work to third parties. This approach eventually ran into problems, as the theory developed in the Court’s rulings did not properly match the facts to which it was applied. So, as the AG observed in GS Media, a hyperlink to a work is not in fact indispensable to the making of that work available on the Internet, but merely “facilitates” locating it online. [70] Eventually, in Ziggo, the Court adjusted the language to accommodate this reality. The current standard is therefore that of “interventions in the absence of which the public would be able to enjoy the work only with difficulty.” [71]

40

The result has been an expansion of the notion of an act of communication to the public to cover not only actual transmissions of works, but also acts that facilitate such transmissions. Following this model, the “conduct element” of accessory copyright liability becomes a strikingly commodious one. The bar is set very low, incorporating any non-minimal participation in the copyright infringement of another party. As the hyperlinking case law reveals, such participation may even take the form of an ex post promotion of the infringing behaviour. The Court appears to place causality at centre stage, so that the existence of some sort of causal connection to the primary infringement, or at least its consequences for the right-holder, seems to be necessary. However, the connection need not be sine qua non. All that is required is that, without the defendant’s involvement, “in principle” [72] infringement would be “more complex”.  [73] The implication is that of a de minimis rule. [74]

41

It is worth noting that, in their national law, EU Member States have often taken different approaches. For example, in English law the doctrine of joint tortfeasance recognises three possible “participation links” connecting the behaviour of parties jointly liable for the same tort: “authorisation”, “inducement” or “procurement”, and entering into a “common design”. [75] As copyright infringement is a tort, the same principles apply to identify accessories in copyright cases. [76] English doctrine emphasises the distinction of these conduct elements from the concept of a “mere facilitation”, which is seen as incapable of giving rise to joint tortfeasance. Similarly, in Germany, under the rule of Article 830(2) of the Bürgerliches Gesetzbuch (German Civil Code – BGB), instigators and accessories to a tort (including copyright infringement) are considered Mittäter, i.e. joint tortfeasors responsible for the damage.

42

Yet, commentators have questioned the conceptual soundness of these approaches. For example, Sales has worried that: “In many cases, […] the dividing line between inducing a third party to commit a wrong and assisting him in its commission is so fine as to be non-existent”. [77] Indeed, careful examination of these notions reveals embedded within them the implication of a mental identification of the collaborator with the primary wrongdoer. Arguably, the terms “authorisation”, “procurement” or “common design” imply an inherent mental dimension. Detangling them from that dimension distorts their natural meaning. Similarly, German legal theory justifies the liability of Mittäter on the existence of a “psychisch vermittelte Kausalität”, i.e. a psychological causation that connects the accessory to the act of the primary infringer. [78]

43

In our view, if it is to be the defendant’s state of mind that determines which conduct leads to accessory liability, then this should be explicitly acknowledged by the law. The mental element should be examined as such, not subsumed within the analysis of the conduct element. Obviously, this approach does not offer great scope for meaningful distinctions between different kinds of intermediary conduct. But this is not necessarily inappropriate. Instead, we suggest that the purpose of the conduct element is to separate the chaff of random bystanders from the grain of the involved accessories, i.e. persons whose liability seems plausible at first glance. The “facilitation” of infringement should accordingly suffice. As all Internet intermediaries will meet this requirement as soon as their services are used by another to commit a copyright infringement, accessory liability will depend on how an intermediary fares in the subsequent analysis. [79]

3.1.2. The Mental Element of Accessory Liability

44

While accessory liability for a copyright infringement cannot be established without a finding of a conduct element, that alone should not be sufficient. Instead, the case law of the CJEU has emphasised that a second condition must be met: that of a mental element. So far, the mental element has only been mentioned by the CJEU in accessory liability cases. As we have already observed, it is reasonable to assume that future decisions will also limit its applicability to such cases, preserving the strict nature of primary copyright liability. The result would be a divide between classic “transmission” cases, for which no mental element is necessary, and “intervention short of transmission” cases, where a mental element would be required.

45

The next question concerns the appropriate threshold for the mental element. General legal theory teaches that two main possibilities exist; namely, the mental element may take the form of intent or knowledge. [80] Again, the CJEU’s case law leads the way. GS Media, Filmspeler and Ziggo all indicate that knowledge of the primary infringement should be sufficient. At the same time, the Court is not oblivious to intent either. Close examination of the wording in Filmspeler and Ziggo makes clear that both concern accessories whose involvement in the infringement was intentional. In the first of these, the Court makes much of the fact that the advertisements of the multimedia player specifically stated that it made it possible to watch audiovisual content made available on the Internet without right-holder authorisation. [81] Likewise, in the second, the Court emphasised the operators’ “purpose to make protected works available to the users” and to “encourage the latter to make copies of those works”. [82] The final conclusion makes sense: while the stronger mental element of intention will certainly suffice, it does not appear to be necessary for accessory liability.

46

A bigger question concerns the type of knowledge required. From the outset, it is important to note that the knowledge relevant to the mental element in accessory liability is knowledge of the lack of authorisation for the primary restricted act. This should not be confused with a condition of volitional engagement in the conduct. Even before GS Media, the CJEU has repeatedly made reference to the fact that the “intervention” representing the act of communication must be made “in full knowledge of the circumstances”, i.e. should be “deliberate”. [83] This wording implies an understanding on the part of the defendant that their conduct is occurring rather than of the infringing nature of the behaviour of another. Its purpose seems to be to exclude inadvertent behaviour from giving rise to liability. [84] It thus appears that, according to the CJEU, deliberately engaging in an “intervention” should be necessary for findings of both accessory and primary copyright infringement. [85] By contrast, the mental element of accessory liability used in GS Media, Filmspeler and Ziggo is clearly focused on the infringing nature of the material that is the object of the “intervention”.

47

In addition to this, two distinctions are important: that between “actual” and “constructive” knowledge and that between “general” and “specific” knowledge. Here, the case law paints a more complicated picture. On the one hand, GS Media is unambiguous. It required that the provider of the hyperlink “knew or ought to have known that the hyperlink he posted provides access to a work illegally placed on the internet”. [86] This points towards a standard of constructive, but specific knowledge, i.e. one that requires that the accessory at least had reason to believe that the specific work to which it was linking was infringing. The rebuttable presumption of knowledge introduced for profit-seeking hyperlinkers reinforces this threshold. By contrast, in Ziggo, general knowledge that the defendant’s services were used to provide access to works published without authorisation from the right-holders was accepted as sufficient. [87] The Court also listed a series of indications of both actual (e.g. notification of infringement) and constructive knowledge (e.g. the large number of torrent files), [88] but was not clear which of these marked the threshold.

48

On the basis of these decisions, we can conclude that both specific and general knowledge should be accepted as relevant. Historically, national European courts have tended to opt for the stricter “specific” approach; however, with the rise of modern technologies, a relaxing of the standard towards “general” knowledge is discernible. [89] Indeed, as the case law shows, the specific or general nature of the knowledge will largely be determined by the underlying conduct element. Whereas in GS Media the provider of the hyperlink is simply “intervening to provide access” to an unauthorised copy of a single protected work, the impact of the management of an online sharing platform is much more severe. Where a broad definition of the conduct element is accepted, a matching mental element is appropriate. Following the GS Media lead, a rebuttable presumption of knowledge may also be introduced in cases where the platform acted in pursuit of a profit. This should only be the case however where the profit-pursuing actions of the platform were targeted at a specific infringement. Simply operating a platform for profit should not be enough. [90]

49

Similarly, given the oft-referenced objective of EU copyright law to provide right-holders with a “high level of protection”, [91] our proposal suggests that both actual and constructive knowledge should be sufficient. This would also be consistent with the general rule applicable in negligence liability in most jurisdictions. [92] At the same time, the wording “ought to have known” carries significant meaning that should not be overlooked: the accessory cannot be expected to go to unreasonable lengths to uncover infringements it has no reason to suspect. Again, the case law of the CJEU is helpful. In L’Oréal v eBay, the Court relied on the concept of a “diligent economic operator” to determine when a host service provider should be considered to have “awareness of facts or circumstances from which the illegal activity or information is apparent” under Article 14 ECD. [93] The same notion can be re-purposed for the construction of a harmonised regime for accessory copyright liability. Thus, the intermediary should only be considered to have constructive knowledge when it is aware of facts or circumstances from which the infringement would have been apparent to a “diligent economic operator”. Reliable notification should suffice as an indication of actual knowledge.

50

Most importantly, the general monitoring prohibition of Article 15 ECD must be respected, as must the limits set by the Charter which underpin this prohibition. An accessory cannot be said to have constructive knowledge where that knowledge could only be acquired through monitoring all of the information it handles. Up until now, these limits have only been considered by the Court in the context of injunctions, but there is no reason why they should not apply more generally: if the imposition of an obligation on an intermediary would disrupt the “fair balance” between fundamental rights even where this is ordered by a court, that should all the more be the case when imposed ex ante with no judicial oversight.

3.1.3. The Violation of a Reasonable Duty of Care

51

The above expansive conclusions on the scope of the mental element do not mean that it has no practical relevance. We propose that the qualification of the mental element have an impact on the final piece of the puzzle: the consequences of the accumulation of a conduct and mental element for the defendant. In this area too, the case law of the CJEU leads the way. To date, outcomes have differed from decision to decision.

52

In GS Media, the Court noted that right-holders have the possibility of informing the providers of hyperlinks that the work to which they have linked has been communicated to the public without permission and of taking action against them if they refuse to remove the links. [94] Whether notification represents a hard condition binding right-holders before a hyperlink provider can be found liable, is not made clear by the decision’s wording. If so, the result would be the establishment of a NTD regime. [95] In Filmspeler and Ziggo, by contrast, no mention of an opportunity to avoid liability through take-down is made. The Court simply concludes that where the conduct element is accompanied by the appropriate mental element, a communication to the public has taken place. [96]

53

We suggest that the explanation to this apparent inconsistency can be found in the nature of the mental element. As noted above, in Filmspeler and Ziggo, in addition to mere knowledge, there were also indications of intention to promote infringement on the part of end-users. It is arguable that these drive the harsh outcome for the intermediaries in those decisions. By contrast, in GS Media, the only possibility considered was that of knowledge. [97] This is a reasonable interpretation that would point towards a sophisticated liability scheme. Indeed, attaching different consequences to different degrees of mental involvement in the commission of a tort has strong roots in European tort law. [98] We suggest that this approach should be further pursued in EU accessory copyright liability. A sensible framework could require that, if an accessory intended an infringement, its behaviour must be understood to be, by definition, unacceptable. Liability should therefore automatically ensue. On the other hand, if the intermediary has no intent, but does have knowledge of the infringement, the violation of a duty of care must first be established before liability can be found. [99]

54

What is missing in the CJEU decisions to date is any consideration of what is to happen where the defendant had only general knowledge (e.g. through notification or due to large amounts of infringing content) that it was facilitating infringement. What we suggest is that the duties of care to be imposed on an accessory should be tailored to the kind of knowledge they display: the type of duty should depend on the type of knowledge.

55

Indeed, it is appropriate that, as GS Media suggests, if the intermediary has specific knowledge – whether actual (e.g. through notification) or constructive (e.g. due to circumstances that would have made it apparent to a “diligent economic operator”) – of a particular infringement it has through its conduct supported, it will often be reasonable to expect that it remove the infringing content. We suggest that, depending on the circumstances, other measures (including preventive ones) might also be appropriate, e.g. the suspension of repeat infringers, notifying the authorities or the provision of identifying data on the perpetrator to the authorities. [100] On the other hand, if the intermediary has only general knowledge of mass infringements taking place using its systems, the removal of content will become much harder. Without specific knowledge, the intermediary will have difficulties locating the infringement without both violating its users’ fundamental rights and expending disproportionate time and energy (thus threatening its own freedom to conduct a business). Other measures may however be considered. The posting of warning duties is an obvious candidate. In all cases, in view of the general monitoring prohibition of Article 15 ECD and the limits set by the Charter, accessories should not be expected to proactively search for infringing content. As a result, notice-and-stay-down regimes should be excluded. All duties of care imposed on accessories should be “reasonable”. [101]

56

Where the accessory fails to take the measures due, assuming it had the ability to take those measures or at least ought to have ensured that it had that ability, it should be held liable. Liability may be for monetary compensation, injunctive relief or both, depending on policy considerations. In our view, no liability should be imposed if the accessory: (a) had no knowledge; (b) having knowledge, took all measures appropriate to the knowledge it possessed; or (c) had knowledge, but no ability to take appropriate measures nor an obligation to ensure that it had such an ability. A final possibility may also be considered: if the intermediary did not take the appropriate measures that would have been expected of it on a flagrantly persistent basis, intent can arguably be inferred and full liability for damages applied. [102]

3.1.4. Interim conclusion

57

As the above analysis reveals, the basic elements for a harmonised European system of accessory liability for copyright and related rights infringement can already be detected in the case law. What is missing is their organisation into a self-standing structure and the application of that structure on a broad basis. At the moment, accessory liability is folded into the rules of primary liability. This is, moreover, only the case for communication to the public: no accessory liability regime for participation in the infringement of other exclusive rights is established. Rather than cramming accessory liability into a single exclusive right, a clear framework distinguishing it from primary liability should be introduced. Ideally, this framework ought to be set by the legislator in a systematic way, instead of trickling out of the CJEU on a case-by-case basis.

58

Of course, a more targeted approach could also be envisaged, tailored specifically to Internet intermediaries or certain types thereof (such as OCSSPs). The preference should however be for a more generalised harmonisation that can allow for a future-proof regime, reliant on principles rather than reactive to circumstances and the pressure of lobby groups. Moreover, unlike Article 17, any new provision should complement, rather than undermine the pre-existing framework and its CJEU interpretation.

59

Would this system alone give right-holders the tools to eliminate online copyright infringement? Undoubtedly, the answer is “no”. A properly balanced regime for accessory copyright liability – particularly one that respects fundamental rights – would not hold intermediaries liable for online infringements where there is no mental element or where duties of care have not been violated. It therefore can only offer part of the answer. Once it is in place, both intermediaries and right-holders will have a clear idea of their legal standing. If no liability can be imposed on the intermediary, attention should then shift to a different question: how should the primary liability of the end-users be addressed?

3.2. An Alternative Compensation System for Content-sharing Platforms

60

The fragementation in the area of intermediary liability can be dealt with through European harmonisation. By contrast, online infringements by the end-users of content-sharing platforms requires more innovative solutions. One such solution can be found in the adoption of an “alternative compensation system”. Generally speaking, this term refers to legal models that replace direct authorisation of certain types of online activities with a scheme for licensing such use and ensuring remuneration to right-holders or at least individual creators, i.e. authors and performers. [103] The use in question would be “permitted but paid”. [104] Different legal models could in theory support such a scheme, ranging from voluntary, extended and mandatory collective management to compulsory or statutory licences. These models impose increasing levels of restriction on exclusive rights. So-called “voluntary” models merely affect the exercise of the right, whereas “mandatory” models impact its very nature, replacing an exclusive right for certain uses with a corresponding right of remuneration or fair compensation. [105] These models do not seek further to strengthen the enforcement measures available to right-holders. Rather, they take a broader normative view of the copyright system as one that should enhance lawful access to and use of works, while ensuring fair remuneration for right-holders, in particular creators.

61

One collective licensing mechanism that has received attention as a possible solution for online uses is that of collective licensing with an extended effect. [106] The DSM Directive significantly harmonises this authorisation mechanism in its Article 12, [107] which envisages three different mechanisms that have the effect of extending collective licenses to non-represented right-holders and their works in a certain territory: extended collective licensing, legal mandates, and presumptions of representation. [108] Article 12 subjects collective licensing with an extended effect to a number of requirements: the licensing mechanism must be managed by a CMO, within well-defined areas of use, where direct licensing is too costly or impractical, and in a manner that “safeguards the legitimate interests” of right-holders. [109] The provision further sets out a number of safeguards that must be put in place for this mechanism to be valid: the CMO must be sufficiently representative; there must be equal treatment between represented and non-represented right-holders; non-represented right-holders must be able to opt-out of the system easily and effectively at any time; and CMOs have to comply with several information obligations towards right-holders. [110]

62

Although there is some debate as to the legal nature of this mechanism, we are of the view that collective licensing with extended effect is not a copyright exception or limitation, [111] at least where the above safeguards, and in particular an opt-out, are put in place. Instead, using the terminology above, this would be a “voluntary” model. [112]

63

The mechanism of Article 12 has been viewed as a possible option for obtaining the authorisations for OCSSPs required by Article 17 of the DSM Directive. Arguably, it could also provide a promising solution in the context of our proposed model. Still, as discussed elsewhere, it faces significant challenges. [113] To begin with, Article 17 does not create a framework for pan-European licensing, but solely for voluntary territorial collective licenses. In addition, it sets out a number of requirements that will be challenging to meet in the short and medium-term for CMOs in many Member States, such as that of representativeness. Even for Member States where some CMOs in some fields meet the requirements (e.g. online music), other CMOs in other fields may not (e.g. visual arts). The result would be the co-existence of extended collective licenses in some Member States (of parts thereof) with a system of individual or limited voluntary collective licensing for platforms in other Member States, leading to a fragmented EU landscape. Finally, because extended collective licensing is an alien mechanism for many Member States, its roll out will necessarily be slow and complex, making it more of a long-term proposition, even if it is adopted consistently across the EU for all types of uses by content-sharing platforms.

64

In light of the above, we are of the view that for types of mass use where enforcement is ineffective, too costly, or even undesirable – due to conflicts with fundamental rights and other public interest concerns – mandatory variants of alternative compensation systems are compelling propositions. Indeed, such systems are best suited to deal with the substantive and territorial fragmentation of copyright that characterises the copyright acquis, to provide legal certainty, and to ensure the participation of users and right-holders in the system. [114]

65

Building on precedents with a long tradition in copyright law – such as continental European private copying schemes – a number of authors and policy makers have been calling for similar solutions in the current “value gap” debate, as supplements to the harmonisation of intermediary liability at the EU level. [115] Our proposal goes in a similar direction. In essence, it aims to legalise and remunerate a broad swath of infringing online acts by individuals (discussed in 2.2.) that are carried out through the services of certain Internet intermediaries. Combined with the accessory copyright liability regime proposed in 3.1., our alternative system could close the regulatory “value gap”, which arises from the lack of adequate rules to deal with mass uncompensated online infringements by users in a way that is consistent with the goals of copyright law and achieves a “fair balance” between competing rights and interests. The scope and main elements of the proposal are outlined below.

3.2.1. Scope of the system

66

The scope of our proposed system includes different design elements. Namely, it is a statutory license coupled with a mandatory exception, applying to specific types of subject matter and specific activities of users (and, by extension, of platforms), coupled with a right of fair compensation, and accompaned by certain obligations on payment and safeguards for the affected platforms. These elements are elaborated on below.

3.2.1.1. Statutory licence and mandatory exception

67

The system we envisage would involve a statutory licence, based on an exception for individual online users. [116] The exception would privilege the non-commercial use of works on user-upload or user-generated content platforms, jointly referred here as “content-sharing platforms”. Such platforms would include Internet intermediaries that provide hosting services enabling users to upload and exchange inter alia copyright-protected works and related subject matter.

68

The exception should be mandatory, not only because of the cross-border and single market relevance of the uses covered, but also due to its strong normative underpinnings. As argued elsewhere, an exception of this type – i.e. one applying to user-generated content – will improve legal certainty for users and content-sharing platforms. [117] It will furthermore align copyright with online social norms, resulting in benefits for the respect and legitimacy of the law. Finally, an exception that covers user-generated content would function as an online extension of the concepts of quotation, parody and transformative use, which are clearly grounded in freedom of expression and information. [118] In its case law, the Court has used this fundamental freedom to limit remedies for copyright infringement and to justify the existence and broad interpretation of the parody and quotation exceptions. [119] In view of this, an optional exception would risk insufficient harmonisation and fail to protect adequately the fundamental rights dimension of privileged uses. [120]

69

The exception would directly cover and authorise acts by individual end-users of content-sharing platforms that are natural persons. This means that neither legal persons nor other collective entities would be direct beneficiaries of the system. This marks a difference from the DSM Directive’s approach. According to Article 17(2) DSMD, if OCSSPs obtain a licence from right-holders, this must also extend to acts of users of their services covered by the right of communication to the public, provided they “are not acting on a commercial basis or where their activity does not generate significant revenues”. Hence, where such agreements are concluded, the provider will assume liability for infringement by end-users by virtue of its contractual arrangement with right-holders. In a way, the platform will be passing on its licence to its users for their non-commercial activities. [121] Where no licensing agreements are concluded by the platform, end-users will not be shielded from liability – unless their activities are privileged by an existing exception. [122] In this no-agreement scenario, platforms will only be able to escape liability if they can show they have made the “best efforts” to license required by Art. 17(4) DSMD. [123]

70

By contrast, our proposal does not rely on the licensing of exclusive rights through platforms. Rather, it focuses on an exception to the benefit of users. At the same time, it would indirectly benefit certain content-sharing platforms, on the condition that they are compliant with the accessory copyright liability regime sketched out above. In practice, this would mean the following.

71

To qualify, a platform must not be making an own use of the work that triggers the exclusive rights of reproduction and communication or making available to the public under Articles 2 and 3 of the InfoSoc Directive. [124] If there is no own use, whether the content-sharing platform is accessorily liable should be considered, meaning that it is necessary to examine its conduct and mental elements. [125] By exempting the acts of individual users, this regime would heighten the threshold required for a finding of knowledge or intent of infringement by the platform as regards uses outside the scope of the exception. This is because accessory liability would depend on the platform having knowledge that the use in question is either commercial or otherwise not covered by the exception.

72

For platforms, this solution would bring significant benefits as compared to the approach in the DSM Directive. It would increase legal certainty by clarifying their liability for acts of their users, while preventing the extension of the exclusive right to their normal activities. For acts of end-users covered by the exception, platforms would be allowed significant breathing space to provide their services and would not be subject to injunctions under Article 8(3) of the InfoSoc Directive. Finally, due to the privilege granted to users, our proposal (in contrast with Article 17(4) DSMD) discourages preventive filtering, an aspect that bolsters its compliance with fundamental rights and the ban on general monitoring in Article 15 ECD. At the same time, as specified above in 3.1.3., there remains ample space for reactive duties of care, such as NTD obligations, to be imposed on platforms, upon obtaining knowledge of infringements regarding content or uses outside the exception’s scope.

3.2.1.2. Subject Matter Scope

73

In theory, our system could apply to all types of protected works and other subject matter, domestic or foreign to an EU Member State, susceptible to upload and use in a content-sharing website. In practice, for reasons of compliance with the three-step test, [126] some subject matter exclusions are sensible. As explained below, the exception underlying our proposal operates as a restriction on the exclusive rights of reproduction and communication to the public in Articles 2 and 3 InfoSoc Directive. It is therefore logical to follow the directive’s regime and exclude computer programs and databases from the scope of our statutory licence. [127] This exclusion is further justified by the idiosyncratic motivations and legal regimes for these types of subject matter in EU law, and the fact that the software and database sectors operate according to specific logics, so that they do not suffer from the problem of large-scale infringement to the same extent as other types of content. [128] In our view, many of the same arguments are valid for videogames. [129] A practical consideration strengthens such exclusions: these categories have, to date, “remained largely exempt from statutory or collective licensing”, on which our system relies. [130]

74

In essence, the main policy choice is whether to extend the system and underlying exception to all or most types of works/subject matter covered by the InfoSoc Directive, or advance a category-based approach, aimed solely at certain types of content popular across content-sharing platforms and for which – and to the extent that – there are mature structures of collective rights management: music, including the musical work, fixations of performances and phonograms; text; visual; and film/video. [131] The best option would be to have the system mirror the subject matter scope of the InfoSoc Directive, as that would ensure the internal consistency of the statutory license. The obvious issue with this approach is that sectors with under-developed collective rights management structures would need to develop these before being able to fully benefit from the system.

3.2.1.3. Substantive Rights Scope: Authorised Acts

75

Our proposal is based on an exception for non-commercial online acts of reproduction and communication to the public by users of content-sharing platforms. For the types of use covered by the exception, the exclusive rights at issue would be recast as rights to fair compensation.

76

Existing proposals differ somewhat in this respect. For example, Hilty and Bauer suggest an exception that would cover the act of uploading and any uses in preparation of user-generated content. [132] In a slightly different approach, the proposal by Senftleben et al. also covers the further dissemination of user-generated remixes and mashups of protected content. [133]

77

In our view, the exception must clearly apply to any use covered by the rights of reproduction and communication to the public that takes place in the context of content-sharing platforms. [134] It should further clarify that it applies to transformative uses (e.g. certain types of remixes or mashups), including those that lie in the grey area between reproduction and adaptation. [135] To be sure, this would amount to an indirect (partial) harmonisation of the right of adaptation. Still, it would prevent legal uncertainty in the application of the exception, which could undermine its effectiveness.

78

As a matter of legal design, the application of a statutory licence to at least two exclusive rights can give rise to one or more copyright exceptions. For example, Article 5(2) of the InfoSoc Directive lists cases allowing for a limitation to the reproduction right, whereas Article 5(3) of the same directive, as well as the Orphan Works Directive, set out exceptions to the rights of reproduction and communication to the public. Therefore, in this respect, our exception is not a radical departure from existing law.

79

Still, we should be mindful to avoid overlaps with existing exceptions susceptible to application to the types of online use targeted here. These include temporary and transient copying, quotation, incidental inclusion and caricature, parody, and pastiche. [136] In fact, Article 17(7) of the DSM Directive imposes an obligation on Member States to ensure that users of OCSSPs benefit from certain exceptions defined as optional in Article 5 InfoSoc Directive. The exact legal nature and scope of the exceptions in paragraph (7) are unclear at this stage. At the very least, the provision will have the effect of making such exceptions mandatory and fully applicable in the context of uploads in OCSSPs. In any case, while potential overlaps cannot be excluded, their existence is not detrimental to our proposal. After all, overlaps already exist in the InfoSoc Directive, for example between private copying and reprography. [137] What is important is to take such overlaps into consideration when designing the system and calculating the amount of compensation due. [138]

80

The thornier issue lies with the central characteristic of the exception; specifically, that it privileges only non-commercial use. The definition of non-commercial use is still an unsettled matter in the EU copyright acquis, despite the fact that the concept features in various provisions of the legal framework. [139] The concept also lies at the heart of the above-mentioned Article 17(2) DSM Directive. [140]

81

As argued elsewhere, the notion of “non-commercial” should be understood as a legal standard (as opposed to a rule) and an autonomous concept of EU law. [141] In essence, this means that it should apply to the use of works by individuals not in direct competition with use by the copyright-holders. To determine this standard, recourse can be had to criteria that are both subjective, like the profit-making purpose of the user, and objective, such as the commercial character of the use. [142] In the context of content-sharing content platforms, where most individual users do not carry out a business activity or make profit from the platform, the application of such a standard will, as a rule, be straightforward.

82

Other authors follow a similar approach. In a nod to the second condition of the three-step test, [143] Hilty and Bauer propose that whether a use is non-commercial or not, should be determined by excluding uses that conflict with the normal exploitation of works because they are on a commercial scale or serve a commercial objective. [144] To make the distinction clear, the authors propose adding recitals to support the exception, listing positive and negative examples, such as an exclusion of uploads in peer-to-peer platforms. [145] Further recitals could be envisaged to flesh out subjective and objective criteria to access the commercial nature of a use. For example, it could be stated that the concept of non-commercial use focuses on online activities of consumption, enjoyment, reference, or expression, outside the context of a market or business activity, and excludes acts with a direct profit intention or for monetary compensation. [146] It would also be sensible, in line with Article 17(2) DSM Directive, to include in the concept activities that do “not generate significant revenues”, as these are unlikely to encroach upon the normal exploitation of works or produce harm to right-holders.

83

An important point is that the concept does not exclude acts of end-users that, although not carried out with a profit-making intention, nonetheless bear economic significance. To exclude from the exception any type of use or works with economic value would deprive the exception of any useful scope and effect, limiting it to de minimis use. The example that comes to mind is that of cost savings for users that would be willing to pay for access to the work outside the content-sharing platform. The payment of fair compensation would be designed to address such potential harm to right-holders, returning it to tolerable levels, provided the permitted use does not conflict with major sources of right revenues of copyright-holders. [147] Grey area cases on the commercial/non-commercial qualification of a use will be decided by national courts and ultimately by the CJEU, in the interpretation of this autonomous concept. [148]

84

A related aspect is the need to ensure the system enables the commercial exploitation of works on content-sharing platforms. Activities like the automatic posting of advertisements on the same page as the work, but unrelated to its content do not appear problematic to us at the moment. Conversely, the direct monetisation of a work by either the end-user or the platform (e.g. through advertisements on an uploaded video tailored to its contents) should generally fall outside the scope of the exception, as it would amount to a commercial use. This would mean that such monetisation would require an agreement with the right-holder, although intermediaries might avoid liability where there is no agreement, if they abide by the conditions of the applicable intermediary liability regime. If such agreement is reached, the commercialisation of the work must be taken into account when calculating fair compensation for its upload, as it would necessarily mean that, from that moment on, no “harm” would be caused to the rights-holder. The result would be that the work would remain accessible, but no fair compensation would be due under our system, as there would be no harm to compensate: the right-holders would already be remunerated through contract.

85

A key point is that the exception would not require the content-sharing platform at issue to be itself a non-commercial enterprise. The non-commercial restriction only refers to the activities of the user in uploading and disseminating the content through the platform, that is neither making its own use of the content nor qualifies as accessorily liable. This precision is important, as it ensures the effectiveness of the limitation and its potential to generate significant revenue for right-holders as fair compensation, by allowing the system to include for-profit platforms.

86

Finally, Hilty and Bauer also propose that only works that are freely available online (being either uploaded from an authorised source, covered by an exception or limitation or by an implied license) can be privileged by the exception. [149] This requirement mirrors that of “lawfulness of source” developed by the CJEU in ACI Adam when interpreting the private copying limitation in light of the three-step test. [150] Although we have reservations as to whether this requirement is truly necessary under the three-step test, its application in the present context would be useful, as it would provide a clear legal basis for right-holders to notify platforms that are otherwise (prior to this knowledge-making notification) not accessorily liable, so that they may remove or disable access to the infringing copy. It also makes clear that the re-uploading of content available online through commercial channels subject to technical restrictions of access (e.g. via a streaming service like Spotify or Netflix) would not be covered by our proposed exception.

3.2.2. Fair Compensation Right and its Calculation

87

The statutory licence proposed here relies on an exception to exclusive rights tied to a right of equitable remuneration [151] or fair compensation. [152] These concepts are not synonymous. Equitable remuneration rights are commonly tied to compulsory licences for uses or forms of exploitation of works not previously covered by an exclusive right. Examples at the international level are the equitable remuneration rights for the broadcasting of works and for “secondary uses” of phonograms. [153] In the EU acquis, there exists a right for an unwaivable and equitable remuneration for rental, an optional right of remuneration for public lending, and rights for a single equitable remuneration for broadcasting and communication to the public. [154] The CJEU interprets the concept of equitable remuneration as based on the “value of the use in trade” and following a logic of balance between competing interests. [155]

88

“Fair compensation” has different contours. This concept accompanies certain limitation-based statutory licences. Before the InfoSoc Directive, most European systems would subject compensated limitations to the payment of equitable remuneration, following the blueprint of other non-exclusive rights in the acquis. After the directive, that concept was replaced by “fair compensation”. Rights of fair compensation are now recognised for reprographic reproductions, private copying, reproductions of broadcasts by non-commercial social institutions, and for use of orphan works by specific organisations. [156]

89

In our view, the concept of fair compensation appears to be the most suitable companion to the exception. First, Recital 36 of the InfoSoc Directive suggests a preference of the EU legislator for this concept in connection with future compensated limitations on the directive’s exclusive rights, including reproduction and communication to the public, i.e. the rights to which our proposed exception applies. Second, there is a vast body of CJEU judgments interpreting “fair compensation” as an autonomous concept of EU law, thus providing a solid basis for its seamless implementation in the acquis. [157]

90

To be consistent with EU law, our exception should be aligned with this autonomous concept, as interpreted by the CJEU. Under this case law, the right to fair compensation is defined as unwaivable and vesting solely in the authors and related right-holders that in EU law are granted the exclusive rights affected by the exception, i.e. those listed in Articles 2 and 3 of the InfoSoc Directive. [158] As a result, the grant of a right of fair compensation ensures, first, that creators receive a share of the amounts collected under the statutory licence system and, second, that they are not forced to transfer that share to exploiters, i.e. publishers and other derivative right-holders. [159]

91

Since the normative concern with the remuneration of creators is at the heart of the “value gap” debate, it makes sense to adopt a fair compensation right that is both unwaivable and primarily for the benefit of creators. This would ensure a steady flow of rights revenue to authors and performers, thereby incentivising and rewarding creative efforts, while protecting creators from ill-advised transfers of their rights to other parties in the context of unbalanced contractual negotiations. [160]

92

The amount of fair compensation should reflect the harm suffered by creators and other right-holders of the affected exclusive rights. [161] The potential harm caused by the introduction of an exception in the context of our alternative compensation system should be considered in terms of the prejudice suffered by copyright owners due to their inability to exercise their copyright for the non-commercial online use of works.

93

Normally, harm would be measured against the effects of the limitation on the market. But there is no actual market to determine accurately the price of the non-commercial use in question. A statutory licence of this type applies to public goods or goods not yet released on the market, [162] and to forms of use for which there is no clear market due to the lack of exercise of the right or its monetisation. [163] Therefore, we suggest that the harm should be calculated on the basis of a different reference point, namely by assessing the value of the uses at issue to the end-users that benefit from the proposed system.

94

Following our previous research, the harm in question could be determined by measuring users’ willingness to pay for such a system through methods of contingent valuation. [164] Furthermore, the amount of harm should take into consideration mitigating factors already referred in the InfoSoc Directive’s recitals and CJEU case law in this field: the de minimis nature of a use, prior payments for such use, or the application of technological protection measures. For instance, certain types of use permitted by the exception only cause minimal harm or are already compensated for through contractual means, i.e. priced into the purchase or licence of the digital content uploaded to the content-sharing platform. The result is that such uses do not cause economic harm to the copyright holders and may not give rise to an obligation to pay fair compensation. [165]

95

Naturally, these factors can be developed and additional ones can be devised to assist in the calculation of the compensation. An example of an existing factor that can be adjusted to our proposal would be to consider the above-mentioned case of post-upload monetisation of content on the platform with the authorisation of right-holders as harm-reducing. Additional factors suggested by other authors include the nature and operation of a platform, as well as the period of time since the publication of uploaded works. [166]

96

Finally, we should reiterate that the scope of our proposed exception is linked to that of existing unremunerated exceptions, such as quotation, criticism, review, caricature, parody or pastiche. [167] If an upload by a user falls within the scope of such exceptions, as interpreted by the CJEU, then it will by definition give rise to no harm to be compensated under our proposed system.

97

If our proposal were to be accepted, it would bring with it consequences for the legal admissibility of the underlying exception. Harm calculated in the way we suggest would correspond to the “prejudice” suffered by copyright-holders due to their inability to exercise and license their rights for the uses under appreciation. By securing the payment of such an amount as fair compensation, the prejudice would be returned to reasonable levels. In so designing the system, we aim to ensure that the exception meets the requirements imposed by the final condition of the three-step test. [168] This approach would also retain the required link between harm and privileged use, while approximating fair compensation to the notion of “appropriate reward” to incentivise creativity, thereby fulfilling a declared objective of EU copyright law. [169]

3.2.3. Payment Obligations and Safeguards for Platforms

98

Finally, it is possible to identify some high-level features of the basic operation of the proposed statutory licence, including payment obligations and safeguards for platforms. The obligation to pay compensation would lie with content-sharing platform providers whose users benefit from the exception proposed. These providers would be considered “paying agents” or debtors in the context of a statutory licensing system managed by a CMO. [170] There are valid policy arguments, linked to the promotion of technological development, that justify an exclusion for certain smaller, younger and/or less profitable platforms from the payment obligation. [171] In our view, subject to an impact assessment, it would be justified to consider payment exemptions for at least two types of service providers. First, for providers that are start-ups, meaning that their services have not been available to the public in the Union for long (e.g. only three or five years). This exemption should operate irrespective of the platform’s size, user-base and/or audience during the period in question. Second, irrespective of their age, an exemption should apply to service providers with an annual turnover below a certain threshold. [172]

99

The statutory licence would be tied to obligatory collective management of the fair compensation right. [173] Subject to the rules set out in the CRM Directive and implementing legislation, cross-border articulation of CMOs could rely on a network of representation agreements of national members to manage and enforce the fair compensation right. It is also conceivable that an EU-wide umbrella CMO centralises certain aspects of the system. [174] The management of the system and the number of organisations involved would depend on the scope and design of the exception(s). As noted, the most efficient design would be that of one exception covering two exclusive rights – reproduction and communication to the public – while leaving existing exceptions in the acquis intact, and dealing with potential overlaps when calculating fair compensation. [175]

100

Collection of the rights revenue should be carried out by the competent national CMO directly from the content-sharing platform. As regards distribution of compensation, the beneficiaries would be the holders of the exclusive rights affected by the exception. Because the right of fair compensation would be unwaivable and non-transferable, authors and performers would be entitled to a fair share of the distributable rights revenue, with the remaining amount going to other right-holders. In our view, creators should be apportioned at least half of the rights revenue collected. [176]

101

As noted, our proposal relies on a functioning and mature system of collective rights management. [177] Moreover, such a system would require adaptation to the digital environment so as to apply to content-sharing platforms. There is therefore an important policy choice to be made as to the scope of application of the system that will directly affect the functioning and management of the statutory licence. If the objective would be to enable the circulation of works and incentivise collectivised licensing, a possible approach would be to accept a broad subject-matter scope for the exception, but to subject payment obligations to the existence of a functioning and sufficiently representative CMO in that area, in the respective Member State.

102

As noted, content-sharing platforms are the primary targets of the compensation, however this would lie ultimately with the users who benefit from the exception. As with most levy systems, payee platforms would have the option to either shift the burden of the compensation to users (e.g. as a subscription fee) [178] or absorb (part of) that cost, for example by financing it out of advertising revenue. [179] Where the payment of compensation is passed on, platforms should have an obligation to make this transparent to users.

103

The imposition of the payment obligation should be counterbalanced by the introduction of safeguards for content-sharing platforms. We suggest that, following the considerations set out above, the new regime should clarify and strengthen the prohibition on the imposition of general monitoring obligations set out in Article 15 ECD. Indeed, if a platform does not intend or have knowledge of the infringement; having knowledge, does not violate any applicable duty of care; and complies with its payment obligations under the proposed statutory licence, it should not be subject to preventive obligations regarding the content it hosts. In such a context, platforms should only be subject to obligations to take action against infringing content where it can be shown that they intend to cause infringement or after obtaining knowledge of an upload that is in violation of the exception. In a world where most of the content uploaded by users would be privileged by an exception, such obligations would become less burdensome for platforms.

3.2.4. Outcome: A Fair, Effective and Future-Proof System

104

Section 3.2. argues for the adoption of a statutory licence tied to a mandatory exception for individual online users. The exception would cover non-commercial use – reproduction and communication to the public (including transformative use) – of works and other subject matter (with the possible exception of software, databases and videogames) previously uploaded legally on content-sharing platforms, whether or not these are for-profit. The system would encompass the vast majority of end-user activities that normally take place on these platforms. The exception would be coupled with an unwaivable right of fair compensation to the benefit of creators and other right-holders, with a fair and appropriate share of rights revenue to be distributed to authors and performers. The obligation to pay compensation would lie with the (non-exempted) platform providers whose users benefit from the exception. This would happen through a statutory licensing system managed by a CMO. In return, the legal system would be clarified so as to avoid the imposition of monitoring and filtering obligations on such platforms and their users. The design of the system and the fair compensation right tied to the exception would ensure its compatibility with the three-step test, at least if the test is interpreted flexibly. [180]

105

This solution would serve as a complement to the harmonised knowledge/intention-based accessory liability regime for copyright sketched out in Section 3.1. It would thus step in to ensure income for creators where the rules of liability reach their practical or legal limits. In this way, providers would be called upon to contribute financially to the creativity displayed on their platforms. Conversely, where the new exception would not apply – most obviously, in cases of commercial use of protected works – or where the platform refuses to cooperate, the liability system would intervene. No fair compensation would be generated, but right-holders would be able to exercise their exclusive rights against those engaging in infringement – including, where their conduct and mental involvement justifies this, Internet intermediaries. If intermediaries cannot be held liable themselves, the system would ensure that, following notification of the infringement, they may be obliged to take proportionate action to remove or prevent the infringement.

106

The anticipated effect of the combined adoption of these two mechanisms would allow breathing space for online platforms to conduct their business and innovate, while also adding a novel revenue stream for right-holders – especially for creators – and focusing enforcement measures on bad-faith platforms that foster large-scale infringement. Users and the public in general would benefit from a rich and diverse online ecosystem, where the risks of content censorship and enforcement in the private sphere are diminished, but possibilities for online enjoyment, expression and creation are promoted. Our proposal could thus bridge the “value gap” in a way that is consistent with the objectives of copyright law and respectful of fundamental rights.

4. Conclusion

107

Current EU copyright law is broken, at least insofar as it applies to mass scale online use. In an attempt to update and fix the legal framework, the new DSM Directive includes a so-called “value gap” provision in Article 17, aimed at enhancing the responsibility of certain user-upload content-sharing platforms. This provision imposes obligations that risk incompatibility with existing law and the Charter of Fundamental Rights of the EU. The result, we fear, will be increased legal uncertainty and a legal regime that is more prone to chilling, rather than fostering, innovation. We argue that a fundamental problem with Article 17 is that is misunderstands the real challenge with which EU copyright law struggles, i.e. the proliferation of copyright infringement online in general, resulting from highly complex and fragmented rules on both primary and accessory copyright infringement.

108

There are no perfect solutions to this challenge. Any new proposal will have to be built on top of a highly complex legal framework. Its benefits and drawbacks should be measured not against an ideal system, but the current “value gap” provision and its potential impact on the acquis.

109

In light of these constraints, we advance a pragmatic alternative for fixing EU copyright law through the parallel implementation of two legal mechanisms. The first consists of the adoption of a harmonised EU framework for accessory liability for third party copyright and related rights infringement, capable of determining, among others, when an intermediary – including content-sharing platforms – should be held liable for its users’ infringements. The second is an alternative compensation system for right-holders covering non-commercial online use by individuals in the context of user-upload platforms. This entails the adoption of a mandatory exception and fair compensation right in the context of a statutory licensing system.

110

Our solution, like most levy-based systems, undoubtedly represents a “rough justice” response to a real world problem. [181] However, when properly calibrated along the lines suggested in this article, such an approach could contribute to achieving the “fair balance” between the rights and interests of right-holders and users that the CJEU places at the heart of EU copyright law and to which Article 17 only pays lip service. The joint operation of the two proposed mechanisms would increase legal certainty for all stakeholders, enable the development of the information society, and provide fair compensation for right-holders for uses of their works in the online environment.

* By Christina Angelopoulos, Lecturer in Intellectual Property at the Centre for Intellectual Property and Information Law (CIPIL) of the University of Cambridge, cja58@cam.ac.uk .
João Pedro Quintais, Postdoctoral Researcher and Lecturer at the Institute for Information Law (IViR) of the University of Amsterdam, j.p.quintais@uva.nl . The authors are listed in alphabetical order, and contributed equally to the development of ideas and to the writing of the article. ORCID: http://orcid.org/0000-0001-6462-2392 .



[1] Proposal for a Directive of the European Parliament and of the Council on copyright in the Digital Single Market (COM(2016)593).

[2] 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 [hereinafter: “DSM Directive” or “DSMD”]. For a timeline of the EU Copyright reform process, see CREATe, ‘EU Copyright Reform, Timeline of Developments’ www.create.ac.uk/policy-responses/eu-copyright-reform/ accessed 26 April 2019. For an overview of the DSM Directive, see JP Quintais, ‘The New Copyright in the Digital Single Market Directive: A Critical Look’ (24 July 2019) EIPR (forthcoming) http://dx.doi.org/10.2139/ssrn.3424770 .

[3] Art. 2(6) DSMD. The definition contains a non-exhaustive list of exclusions, including: not-for profit online encyclopaedias, not-for-profit educational and scientific repositories, open source software developing and sharing platforms, electronic communication service providers, online marketplaces and certain storage cloud services.

[4] Art. 17(1) DSMD. NB that Recital 64 states – incorrectly in our view – that this provision is a clarification of existing law.

[5] Art. 17(3) DSMD, explicitly excluding the application of Art. 14(1) ECD.

[6] According to Art. 17(2) DSMD, that authorisation will extend to uploading users, provided their activities are non-commercial.

[7] M Senftleben, “Bermuda Triangle – Licensing, Filtering and Privileging User-Generated Content Under the New Directive on Copyright in the Digital Single Market” E.I.P.R. 2019, 41(8), 480, 481.

[8] Art. 17(4) DSMD.

[9] Charter of Fundamental Rights of the European Union, O.J. 2016, C 364/01 [hereinafter: ‘Charter’].

[10] See e.g. S Stalla-Bourdillon et al., ‘A Brief Exegesis of the Proposed Copyright Directive’ (24 November 2016) https://ssrn.com/abstract=2875296 accessed 26 April 2019 and M Senftleben et al., ‘The Recommendation on Measures to Safeguard Fundamental Rights and the Open Internet in the Framework of the EU Copyright Reform’ 40(3) (2018) EIPR 149-163; C Angelopoulos, ‘Study on Online Platforms and the Commission’s New Proposal for a Directive on Copyright in the Digital Single Market’ (January 2017) https://ssrn.com/abstract=2947800 accessed 26 April 2019.

[11] See Art. 15 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) [2000] OJ L 178/1.

[12] See Art. 17(8) and Recital 66 DSMD. Despite this it is not clear how infringing works are to be located without the platform monitoring all posted content. Recital 66 does suggest that “in some cases” the availability of unauthorised content may “only be avoided upon notification of rightholders”. This presumably means that in other cases it will be for the intermediary to find infringements through general checks. For further criticism, see e.g. M Senftleben, “Bermuda Triangle – Licensing, Filtering and Privileging User-Generated Content Under the New Directive on Copyright in the Digital Single Market” E.I.P.R. 2019, 41(8), 480-490 and Communia Association, ‘A Final X-Ray of Article 13: Legislative Wishful Thinking that Will Hurt User Rights’ (COMMUNIA, 5 March 2019) https://www.communia-association.org accessed 26 April 2019.

[13] Case C-70/10, Scarlet Extended (24 November 2011) ECLI:EU:C:2011:771; Case C-360/10, Netlog (16 February 2012) ECLI:EU:C:2012:85, [41]–[43], [48], [50]–[51]; Case C-484/14, McFadden, (15 September 2016) ECLI:EU:C:2016:689. See also C Angelopoulos, ‘Study on Online Platforms and the Commission’s New Proposal for a Directive on Copyright in the Digital Single Market’ (January 2017) https://ssrn.com/abstract=2947800 accessed 26 April 2019.

[14] Art. 17(5) DSMD.

[15] M Senftleben, “Bermuda Triangle – Licensing, Filtering and Privileging User-Generated Content Under the New Directive on Copyright in the Digital Single Market” E.I.P.R. 2019, 41(8), 480, 484-485. See also E Engstrom and N Feamster, ‘The Limits of Filtering: A Look at the Functionality and Shortcoming of Content Detection Tools’ (Engine, March 2017) www.engine.is/the-limits-of-filtering .

[16] Art. 17(8) DSMD. On the ineffectiveness of complaint and redress mechanisms in this context, see J Urban, J Karaganis & B Schofield, ‘Notice and Takedown in Everyday Practice’ (March 2016) UC Berkeley Public Law Research Paper No. 2755628, ssrn.com/abstract=2755628, 57-62.

[17] See Art. 17(4) DSMD.

[18] See Art. 17(9) DSMD.

[19] Consider in this regard, Case C‑360/10, SABAM v Netlog (16 February 2012) ECLI:EU:C:2012:85, [49].

[20] For a distinction between the concepts of fair compensation and (equitable) remuneration in EU copyright law, see infra at 3.2.2.

[21] For further information, see: C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016) and JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017).

[22] Directive 2001/29/EC of the European Parliament and of the Council of 22 May 2001 on the harmonisation of certain aspects of copyright and related rights in the information society [2001] OJ L167/10 [hereinafter InfoSoc Directive].

[23] 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) [2000] OJ L 178/1 [hereinafter ECD].

[24] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 163–164. Leading CJEU cases in this respect include C-5/08, Infopaq I (16 July 2009), ECLI:EU:C:2009:465; C-403/08, Football Association Premier League (4 October 2011), ECLI:EU:C:2011:631; and C-419/13, Art & Allposters (22 January 2015), ECLI:EU:C:2015:27.

[25] See e.g. A Strowel, “Reconstructing the Reproduction and Communication to the Public Rights: How to Align Copyright with its Fundamentals” and S Dusollier, “Realigning Economic Rights with Exploitation of Works: the Control of Authors over the Circulation of Works in the Public Sphere” in PB Hugenholtz (ed.), Copyright Reconstructed: Rethinking Copyright's Economic Rights in a Time of Highly Dynamic Technological and Economic Change (Wolters Kluwer 2018) 204-206 and 166-169.

[26] JP Quintais, ‘Untangling the Hyperlinking Web: In Search of the Online Right of Communication to the Public’ (2018) 21(5-6) J. World Intell. Prop. 385.

[27] See Case C-135/10, SCF (15 March 2012) ECLI:EU:C:2012:140, [76]-[79]; Case C-162/10, Phonographic Performance (Ireland) (15 March 2012) ECLI:EU:C:2012:141, [29].

[28] JP Quintais, JP, ‘Untangling the Hyperlinking Web: In Search of the Online Right of Communication to the Public’ (2018) 21(5-6) J. World Intell. Prop. 385.

[29] See, for all, Case C-117/15, Reha Training (31 May 2016) ECLI:EU:C:2016:379.

[31] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 152-244.

[32] See however Art. 17(7) and Recital 70 DSMD, which suggest that the exceptions for quotation, criticism, review, caricature, parody or pastiche should be made mandatory.

[33] On the general structure and functioning of Art. 5, see L Guibault, ‘Why Cherry-Picking Never Leads to Harmonisation: The Case of the Limitations on Copyright under Directive 2001/29/EC’ 1 (2010) JIPITEC 55.

[34] Case C-5/08, Infopaq I (16 July 2009) ECLI:EU:C:2009:465, [56]–[57]; Case C-145/10, Painer (7 March 2013) ECLI:EU:C:2013:138, [109], [133]; Cases C-403/08 and C-429/08, Football Association Premier League (4 October 2011) ECLI:EU:C:2011:631, [162]; Case C-360/13, Public Relations Consultants Association (5 June 2014) ECLI:EU:C:2014:1195, [23]. See also Janssens (2009) 323 (considering the doctrine of strict interpretation to be a common view point in international norms) and M Senftleben, ‘Breathing Space for Cloud-Based Business Models’ (2013) 4 JIPITEC 93.

[35] PB Hugenholtz and M Senftleben, Fair Use in Europe: In Search of Flexibilities (Institute for Information Law/VU Centre for Law and Governance 2011).

[36] For an extensive analysis see J-P Trialle and others, ‘Study on the Application of Directive 2001/29/EC on Copyright and Related Rights in the Information Society (the “InfoSoc Directive”) (Directorate-General for the Internal Market and Services’ (European Commission, 2013).

[37] See JP Quintais, ‘Rethinking Normal Exploitation: Enabling Online Limitations in EU Copyright Law’ 41 (2017) 6 AMI: Tijdschrift voor Auteurs-, Media- & Informatierecht, 197-205, proposing the introduction of a mandatory and unwaivable limitation for user-generated content. See also M Senftleben, “Bermuda Triangle – Licensing, Filtering and Privileging User-Generated Content Under the New Directive on Copyright in the Digital Single Market” E.I.P.R. 2019, 41(8), 480, 485-490.

[38] On these objectives as interpreted by the CJEU, see the empirical analysis of M Favale and others, ‘Is There an EU Copyright Jurisprudence? An Empirical Analysis of the Workings of the European Court of Justice’ (2016) 79 The Modern L. Rev. 31.

[39] The argument is developed in detail in JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017). See also C Geiger, ‘Challenges for the Enforcement of Copyright in the Online World: Time for a New Approach’ in P Torremans (ed.), Research Handbook on the Cross-Border Enforcement of Intellectual Property (Edward Elgar 2014) 704 and C Geiger, ‘Statutory Licenses as Enabler of Creative Uses’ in RM Hilty and K-C Liu (eds.), Remuneration of Copyright Owners (Springer 2017) 305-327.

[40] The connection between a “high level of protection”, creative incentives and “adequate remuneration” for right-holders is patent in Recitals 9 and 10 InfoSoc Directive. For the use of “fair balance” and “adequate protection” by the CJEU as teleological arguments to limit a broad interpretation of rights in the copyright acquis, see M Favale and others, ‘Is There an EU Copyright Jurisprudence? An Empirical Analysis of the Workings of the European Court of Justice’ (2016) 79 The Modern L. Rev. 31, 65–68. On the disconnect between EU copyright law and social norms, see PB Hugenholtz and M Senftleben, ‘Fair Use in Europe: In Search of Flexibilities’ (Institute for Information Law/VU Centre for Law and Governance 2011) 26-28.

[41] To benefit from the safe harbour, the host must additionally not have had authority or control over the end-user in the posting of the contested information (Art. 14(2) ECD).

[42] Case C‑360/10, SABAM v Netlog (16 February 2012) ECLI:EU:C:2012:85, [34] and Case C-484/14, McFadden, (15 September 2016) ECLI:EU:C:2016:689, [87].

[43] See Case C-275/06, Promusicae (29 January 2008) ECLI:EU:C:2008:54, [68].

[44] Case C-360/10, SABAM v Netlog, (16 February 2012) ECLI:EU:C:2012:85, [25], [52].

[45] No example of more moderate requests concerning hosting providers have come before the CJEU. However, that Court has found that blocking injunctions issued against internet service providers may strike a “fair balance” (Case C-314/12, Telekabel (17 March 2014) ECLI:EU:C:2014:192), as will obligations for WiFi providers to password-protect their network (see Case C-484/14, McFadden (15 September 2016) ECLI:EU:C:2016:689, [100]).

[46] Case C-117/15, Reha Training (21 May 2012) ECLI:EU:C:2016:379, [46], citing Case C-306/05, SGAE (7 December 2006) ECLI:EU:C:2006:764, [42] and Cases C-403/08 and C-429/08, Football Association Premier League (Oct 4, 2011) ECLI:EU:C:2011:631, [195].

[47] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [26].

[48] Case C-160/15, GS Media (8 September 2016) ECLI:EU:C:2016:644, [37] and [48] . The connection between the ‘new public’ and knowledge was made even more explicit in later case law, see Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [47]-[52] and Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [44]-[45].

[49] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456.

[50] J Ginsburg, ‘The Court of Justice of the European Union Creates an EU Law of Liability for Facilitation of Copyright Infringement: Observations on Brein v. Filmspeler [C-527/15] (2017) and Brein v. Ziggo [C-610/15] (2017)’ (22 August 2017) https://ssrn.com/abstract=3024302 accessed 26 April 2019; M Leistner, ‘Is the CJEU Outperforming the Commission?’ (26 November 2017) https://ssrn.com/abstract=3077615 accessed 26 April 2019.

[51] C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), section 1.3.1.

[52] See e.g. L Bently and B Sherman, Intellectual Property Law (5th ed, Oxford University Press 2018) 143.

[53] This conclusion is confirmed by the recent decision in the primary liability case of Renckhoff, in which the Court made no reference to knowledge. See Case C-161/17, Renckhoff (7 August 2018) ECLI:EU:C:2018:634.

[54] Some light is expected to be shed on these matters through the CJEU’s judgements in Case C-682/18, LF v Google, Case C-683/18, Elsevier v Cyando, and C-500/19, Puls 4 TV, currently pending.

[55] See the references cited supra in note 10. See also, commenting on legislative proposals for the DSM Directive: C Angelopoulos, ‘Axel Voss’s JURI Report on Article 13 Would Violate Internet Users’ Fundamental Rights’ (Kluwer Copyright Blog, 29 June 2018) http://copyrightblog.kluweriplaw.com/ , accessed 17 April 2019; V Moscon, ‘The Council Negotiating Position on Article 13: Warning, Creators are also Part of the Game!!!’ (Kluwer Copyright Blog, 4 July 2018) http://copyrightblog.kluweriplaw.com/2 , accessed 17 April 2019.

[56] Opinion of Advocate General Szpunar at [3], Case C-610/15, Stichting Brein v Ziggo (8 February 2017), ECLI:EU:C:2017:99.

[57] See e.g. PS Davies, Accessory Liability (Hart Publishing 2015). For a comparative analysis see C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), section 5.1.1.

[58] See Recital 40 ECD. Indeed, Recital 40, in explaining the purpose of the safe harbours, talks about providers’ “duty to act under certain circumstances”, language which does not reflect the nature of the safe harbours as mere immunities, but which might foreshadow the introduction of substantive harmonised law.

[59] J Ginsburg, ‘The Court of Justice of the European Union Creates an EU Law of Liability for Facilitation of Copyright Infringement: Observations on Brein v. Filmspeler [C-527/15] (2017) and Brein v. Ziggo [C-610/15] (2017)’ (22 August 2017) https://ssrn.com/abstract=3024302 accessed 26 April 2019.

[60] Case C-607/11, ITV Broadcasting (7 March 2013) ECLI:EU:C:2013:147; Case C-325/14, SBS Belgium (15 November 2015) ECLI:EU:C:2015:764; Case C-138/16, AKM (16 March 2017) ECLI:EU:C:2017:218.

[61] Case C-306/05, SGAE (7 December 2006) ECLI:EU:C:2006:764; Case C-136/09, Organismos Sillogikis Diacheirisis… v Divani Akropolis (March 18, 2010) ECLI:EU:C:2010:151; Case C-135/10, SCF (15 March 2012) ECLI:EU:C:2012:140; Case C-162/10, Phonographic Performance (Ireland) (15 March 2012) ECLI:EU:C:2012:141; Case C-117/15, Reha Training (21 May 2012) ECLI:EU:C:2016:379.

[62] Note that in Phonographic Performance (Ireland), even the provision of apparatus and phonograms was qualified as a “transmission”. See Case C-162/10, Phonographic Performance (Ireland) (15 March 2012) ECLI:EU:C:2012:141, [64]-[69].

[63] Case C-527/15, Stichting Brein v Jack Frederik Wullems (26 April 2017) ECLI:EU:C:2017:300.

[64] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [30].

[65] Id. [32].

[66] Id. [34].

[67] Note that to date, the relationship between a “transmission” and “making available” has not been clearly addressed by the Court. While Recital 23 suggests that “making available” should be understood as a form of transmission, the two could also be seen as separate, applying respectively to linear and non-linear (i.e. on-demand) types of communication. See JP Quintais, ‘Untangling the Hyperlinking Web: In Search of the Online Right of Communication to the Public’ (2018) 21(5-6) J. World Intell. Prop. 385.

[68] Case C-160/15, GS Media (8 September, 2016) ECLI:EU:C:2016:644 , [35].

[69] Case C-527/15, Stichting Brein v Jack Frederik Wullems (26 April 2017) ECLI:EU:C:2017:300, [31].

[70] Opinion of Advocate General Wathelet at [54]-[60], Case C-160/15, GS Media (7 April 2016) ECLI:EU:C:2016:221.

[71] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [26].

[72] Case C-527/15, Stichting Brein v Jack Frederik Wullems (26 April 2017) ECLI:EU:C:2017:300, [31].

[73] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [36].

[74] For further discussions on the de minimis rule in accessory liability see: C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), section 5.2.3.; PS Davies, Accessory Liability (Hart Publishing 2015) 37-40; D Cooper, Secondary Liability for Civil Wrongs (unpublished PhD thesis, University of Cambridge 1995) 9; and P Sales, ‘The Tort of Conspiracy and Civil Secondary Liability’ (1990) 49(3) Cambridge L. J. 491, at note 62.

[75] H Carty, ‘Joint Tortfeasance and Assistance Liability’ (1999) 19(4) Legal Studies 489; L Bently and B Sherman, Intellectual Property Law (5th ed, Oxford University Press 2018) 1286-1289.

[76] See e.g. Credit Lyonnais Bank Nederland NV (now Generale Bank Nederland NV) v Export Credits Guarantee Department [1999] C.L.C. 823, 853.

[77] P Sales, ‘The Tort of Conspiracy and Civil Secondary Liability’ (1990) 49(3) Cambridge L. J. 491.

[78] W van Gerven and others, Tort Law (Hart Publishing 2000) 430-432; and H Koziol, Basic Questions of Tort Law from a Germanic Perspective (Jan Sramek Verlag 2012) 140.

[79] It is worth noting that recent academic approaches to accessory liability espouse a similar approach, including in jurisdictions that have traditionally attempted a more discerning definition of conduct capable of resulting in accessory liability. See, e.g., PS Davies, Accessory Liability (Hart Publishing 2015) 39. For a similar analysis by a national judge see Lord Sumption in Sea Shepherd UK v Fish & Fish Ltd [2015] UKSC 10 (4 March 2015), [41].

[80] See C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), sections 4.2. and 5.3, and PS Davies, Accessory Liability (Hart Publishing 2015) 203-209.

[81] Case C-527/15, Stichting Brein v Jack Frederik Wullems (26 April 2017) ECLI:EU:C:2017:300, [50].

[82] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [45]-[46].

[83] Note that under the general principles of European criminal law, intent covers cases where the actor deliberately acts in a manner she knows will almost certainly bring about a result (“dolus indirectus”). See J Blomsma, Mens Rea and Defences in European Criminal Law (Intersentia 2013) 509-526. The Draft Common Frame of Reference on European civil law also includes within its definition of intention instances where the defendant knew that his behaviour could result in the commission of the wrong by the third party and meant to engage in it. See C von Bar et al. (eds), Principles, Definitions and Model Rules of European Private Law: Draft Common Frame of Reference (DCFR), prepared by the Study Group on a European Civil Code and the Research Group on EC private law (Acquis Group), Outline Edition (Sellier 2009) 3266.

[84] On these two different aspects of the mental element, see PS Davies, Accessory Liability (Hart Publishing 2015) 41.

[85] For a different interpretation see J Ginsburg, ‘The Court of Justice of the European Union Creates an EU Law of Liability for Facilitation of Copyright Infringement: Observations on Brein v. Filmspeler [C-527/15] (2017) and Brein v. Ziggo [C-610/15] (2017)’ (22 August 2017) https://ssrn.com/abstract=3024302 accessed 26 April 2019.

[86] Case C-160/15, GS Media (7 April 2016), ECLI:EU:C:2016:221 [49].

[87] Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456, [45].

[88] Id.

[89] C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), section 5.3.2.2.

[90] It is relevant that in Filmspeler and Ziggo, both of which concerned general knowledge, no rebuttable presumption of knowledge based on profit was introduced, the Court instead limiting itself to indicating that profit is an important factor.

[91] See Recital 9 InfoSoc Directive, as well as the case law of the CJEU (e.g. Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456 , [22]). See also M Favale and others, ‘Is There an EU Copyright Jurisprudence? An Empirical Analysis of the Workings of the European Court of Justice’ (2016) 79 The Modern L. Rev. 31, exploring inter alia the role of this recital in the Court’s “activist, harmonising agenda” of EU copyright law.

[92] C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), section 5.3.2.1.

[93] Case C-324/09, L’Oréal v eBay International (12 July 2011) ECLI:EU:C:2011:474, [120]-[124].

[94] Case C-160/15, GS Media (7 April 2016) ECLI:EU:C:2016:221, [53].

[95] Presumably, GS Media’s rebuttable presumption of knowledge would mark the limits of such a system.

[96] Case C-527/15, Stichting Brein v Jack Frederik Wullems (26 April 2017) ECLI:EU:C:2017:300 , [52] and Case C-610/15, Stichting Brein v Ziggo (14 June 2017) ECLI:EU:C:2017:456 , [47].

[97] Although the posting of the hyperlink at issue in GS Media was, based on the facts, undoubtedly itself intentional (i.e. not inadvertent), it is hard to see how the provider of the hyperlink could be said to have intended the prior uploading of the protected content. The chronology is significant here: it is arguably not possible to intend an infringement that occurred before your own action, but it would be possible to intend the infringements performed subsequently by users of your technology. While GS Media is concerned only with knowledge of the unauthorised nature of the linked-to material, Filmspeler focuses on the advertisement to end-users of the multimedia player’s infringing capabilities. Stichting Brein v Ziggo concerns technology that enables unauthorised file exchanges by others.

[98] FH Lawson and B Markesinis, Tortious Liability for Unintentional Harm in the Common Law and the Civil Law (Cambridge University Press 1982) 181. See also Art. 2:102(5) of the Principles of European Tort Law (PETL), developed by the European Group on Tort Law, which assigns greater weight to intentionally committed torts over merely negligence-based ones: “The scope of protection may also be affected by the nature of liability, so that an interest may receive more extensive protection against intentional harm than in other cases.” The PETL are available at: www.egtl.org . See also European Group on Tort Law, Principles of European Tort Law – Text and Commentary (Springer 2005).

[99] C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), sections 5.3 and 5.4.

[100] The inspiration for these suggestions can be found in the CJEU’s decisions on the safe harbours, see Case C-324/09, L’Oréal v eBay International (12 July 2011) ECLI:EU:C:2011:474, [141].

[101] C Angelopoulos, European Intermediary Liability in Copyright: A Tort-Based Analysis (Kluwer Law International 2016), sections 5.4 and 5.5. Cf., on trademark law, F Mostert, ‘Study on Approaches to Online Trademark Infringements’ (World Intellectual Property Organization 2017).

[102] Precedent for a presumption of intent on the basis of the previous persistent violation of duties of care can be found, for example, in German case law. The German lower courts have, in the past, found that the flagrant and consistent violation of duties of care can push liability beyond Störerhaftung, i.e. liability only for injunctive orders, towards full liability as a Gehilfe, i.e. an assistant. See e.g. OLG München, 11 January 2006, 21 O 2793/05, (2006) 3 ZUM 255; OLG Hamburg, 13 May 2013, 5 W 41/13, (2013) 8 MMR 533Hörspiel; LG Frankfurt a. M. , 5 February 2014, 2-06 O 319/13, (2015) 2 ZUM 160 – File-Hosting. See also M Gruenberger & A Dietz, ‘Germany’ in L Bently, International Copyright Law and Practice (Matthew Bender 2017), § 8[1][c][i]; S Jaworski & JB Nordemann, ‘Gehilfenhaftung von Intermediären bei Rechtsverletzungen im Internet’ (2017) 6 GRUR 567; T Hoeren & S Yankova, “The Liability of Internet Intermediaries – The German Perspective” (2012) 43 IIC 501, section C.II.

[103] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017), describing different types of proposals for such systems, initially designed for the legalisation of peer-to-peer file-sharing.

[104] The term is taken from J Ginsburg, ‘Fair Use for Free, or Permitted-but-Paid? (2015) 29 Berkeley Tech. L. J. 1383.

[105] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 85-150.

[106] JP Quintais, ‘The New Copyright in the Digital Single Market Directive: A Critical Look’ (24 July 2019) EIPR (forthcoming) http://dx.doi.org/10.2139/ssrn.3424770 .

[107] Prior to the DSM Directive, it was possible to find specific references to extended collective licensing in Art. 3(2)–(4) Satellite and Cable Directive, as well as in Recitals 18 InfoSoc Directive, 24 Orphan Works Directive, and 12 CRM Directive. In addition, several Member States contain provisions of this type in their national laws, especially the Nordic countries. See also Recital 44 DSMD.

[108] Art. 12(1)(b) DSMD.

[109] Art. 12(2) and Recitals 45-47, and 49 DSMD. The definition of CMO is found in Art. 3(a) CRM Directive.

[110] Art. 12(3) DSMD.

[111] JP Quintais, ‘The New Copyright in the Digital Single Market Directive: A Critical Look’ (24 July 2019) EIPR (forthcoming) http://dx.doi.org/10.2139/ssrn.3424770 . For a more developed argument regarding extended collective licensing with an opt-out, see JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 110–111 (and references cited therein); J Axhamn and L Guibault, “Cross-Border Extended Collective Licensing: A Solution to Online Dissemination of Europe’s Cultural Heritage” Amsterdam Law School Research Paper No. 2012-22 (8 February 2012) https://ssrn.com/abstract=2001347 accessed 21 August 2018; and J Axhamn, “Exceptions, limitations and collective management of rights as vehicles for access to information. In Access To Information And Knowledge: 21st Century Challenges in Intellectual Property and Knowledge Governance” (Edward Elgar 2013) 164–186.

[112] See, in this respect, JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017).

[113] JP Quintais, ‘The New Copyright in the Digital Single Market Directive: A Critical Look’ (24 July 2019) EIPR (forthcoming) http://dx.doi.org/10.2139/ssrn.3424770 .

[114] Id. 85-150, 242-243.

[115] R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019, 99-112. This proposal is endorsed with slight changes by M Senftleben and others, ‘The Recommendation on Measures to Safeguard Fundamental Rights and the Open Internet in the Framework of the EU Copyright Reform’ (2018) 40(3) EIPR 149. See also, V Moscon, ‘The Council Negotiating Position on Article 13: Warning, Creators are also Part of the Game!!!’ (Kluwer Copyright Blog, 4 July 2018) http://copyrightblog.kluweriplaw.com/ , accessed 17 April 2019; J Reda, ‘Julia Reda discusses the current Proposal for a Directive on copyright in the Digital Single Market’ (Kluwer Copyright Blog, 18 June 2018) http://copyrightblog.kluweriplaw.com/2018/06/18/ accessed 26 April 2019. Prior proposals with relevance to this topic include: JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017); T Kreutzer and others, ‘The Berlin Gedankenexperiment on the restructuring of Copyright Law and Author’s Rights’ 7 (2016) JIPITEC 76, para 1; M Leistner, Copyright Law on the Internet in Need of Revision: Hyperlinks, Online Platforms and Aggregators 12 (2) (2017) J Intell. Prop. L. & Practice 136-149; M Leistner and A Metzger, ‘The EU Copyright Package: A Way Out of the Dilemma in Two Stages’ (2017) 48 IIC 381.

[116] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 26: “Statutory licenses restrict the nature of a pre-existing exclusive right through the application of a compensated limitation.” In this article, we use the term ‘exception’ as having the same meaning as ‘limitation’, and with reference to the title of Art. 5 InfoSoc Directive.

[117] JP Quintais, ‘Rethinking Normal Exploitation: Enabling Online Limitations in EU Copyright Law’ 41 (2017) 6 AMI: Tijdschrift voor Auteurs-, Media- & Informatierecht 197-205.

[118] See Art. 11 Charter and Recital 3 InfoSoc Directive. This point is also recognised in the context of the new mandatory exceptions in Article 17(7) DSMD and accompanying recital 70.

[119] Case C-70/10, Scarlet Extended (24 November 2011) ECLI:EU:C:2011:771, [43], [50]–[53]; Case C-360/10, Netlog (16 February 2012) ECLI:EU:C:2012:85, [41]–[43], [48], [50]–[51]; Case C-314/12, UPC Telekabel (27 March 2014) ECLI:EU:C:2014:192, [46]–[47], [55]–[57]; Case C-201/13, Deckmyn (3 September 2012) ECLI:EU:C:2014:2132; Case C-145/10, Eva-Maria Painer (1 December 2011) ECLI:EU:C:2013:138; Case C-469/17, Funke Medien (29 July 2019) ECLI:EU:C:2019:623, [39]-[54]; Case C-476/17, Pelham (29 July 2019) ECLI:EU:C:2019:624, [72]; Case C-516/17, Spiegel Online (29 July 2019) ECLI:EU:C:2019:625, [50]-[59], [75]-[95].

[120] M van Eechoud and others, Harmonising European Copyright Law: The Challenges of Better Law Making (Kluwer Law International 2009) 128–129, identifying these arguments for qualifying limitations as mandatory.

[121] NB other contractual schemes are possible, for example a covenant not to sue by right-holders, an indemnification obligation by the platforms, a contract in favour of third parties (the users) by the platform, a sub-licence, etc.

[122] Such as applicable exceptions in Art. 5(3) InfoSoc Directive and, after national implementation of the DSM Directive, those mentioned in Art. 17(7) thereof.

[123] Art. 17(4) DSMD. See also Art. 17(5) DSMD, identifying the factors to be taken into account in determining the compliance of platforms with their obligations under Art. 17(4) DSMD.

[124] Our position, contrary to Art. 17(1) DSMD, is that a for-profit platform that merely hosts large amounts of copyright-protected works or other protected subject matter uploaded by its users is not automatically carrying out a restricted act under Art. 3 InfoSoc Directive.

[125] See supra 3.1.1.–3.1.2.

[126] More specifically, subject matter restrictions are important to ensure that the exception underlying our system qualifies as a “certain special case” under the first step of the test. See JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 387-388 and references therein.

[127] See Art. 1(2)(a) and (e) InfoSoc Directive.

[128] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 388 and references cited therein.

[129] See, for a detailed argument, JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 387-391.

[130] PB Hugenholtz and JP Quintais, ‘Towards a Universal Right of Remuneration: Legalizing the Non-commercial Online Use of Works’ in PB Hugenholtz (ed), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological Change (Kluwer Law International 2018) 265.

[131] Id. 266.

[132] R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019.

[133] M Senftleben and others, ‘The Recommendation on Measures to Safeguard Fundamental Rights and the Open Internet in the Framework of the EU Copyright Reform’ (2018) 40(3) EIPR 149.

[134] NB Art. 17(7) DSMD extends the exceptions therein to the activities of users “when uploading and making available content …on online content-sharing services”.

[135] On the potential scope of an exception for user-generated content in the acquis, see JP Quintais, ‘Rethinking Normal Exploitation: Enabling Online Limitations in EU Copyright Law’ (2017) 41(6) AMI: Tijdschrift voor Auteurs-, Media- & Informatierecht 197-205, citing the proposals the Internal Market Committees (IMCO) and the Culture Committee (CULT) during the legislative process of the proposed Directive. See European Parliament, ‘Opinion of the Committee on the Internal Market and Consumer Protection for the Committee on Legal Affairs’ (COM(2016)0593 – C8-0383/2016 – 2016/0280(COD)), Rapporteur: Catherine Stihler, 14 June 2017, Amendment 55, new Art. 5b, 36; European Parliament, ‘Opinion of the Committee on Culture and Education for the Committee on Legal Affairs’ (COM(2016)0593 – C8-0383/2016 – 2016/0280(COD)), Rapporteur: Marc Joulaud, 4 September 2017, Amendment 60, Art. 5a (new) on “Use of short extracts and quotations from copyright-protected works or other subject matter in content uploaded by users”. For a user-generated content exception that was made into law, see Sec. 29.21(1) of the Canadian Copyright Act (Copyright Act (R.S.C., 1985, c. C-42)).

[136] Arts 5(1), 5(2)(b), and 5(3)(d), (i) and (k) InfoSoc Directive.

[137] Arts 5(2)(a) and (b) InfoSoc Directive. See in this respect, Case C-572/13, Reprobel (12 November 2015) ECLI:EU:C:2015:750, [32]-[43].

[138] Cf. JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 391-392. See infra 3.2.2.

[139] Art. 5(2)(b) and (e), 5(3)(a) and (b) InfoSoc Directive, as well as Art. 6(5) Orphan Works Directive (Directive 2012/28/EU of the European Parliament and of the Council of 25 October 2012 on certain permitted uses of orphan works [2012] OJ L 299/5) and Art. 5(3) of the Collective Rights Management Directive (Directive 2014/26/EU of the European Parliament and of the Council of 26 February 2014 on collective management of copyright and related rights and multi-territorial licensing of rights in musical works for online use in the internal market [2014] OJ L 84/72).

[140] See supra 3.2.1.1.

[141] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 10-11, 214-215, 276-284, 392-394. See also PB Hugenholtz and JP Quintais, ‘Towards a Universal Right of Remuneration: Legalizing the Non-commercial Online Use of Works’ in PB Hugenholtz (ed), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological Change (Kluwer Law International 2018) 263-264.

[142] Id. 391-394 and references cited therein. NB Art. 17(2) DSMD appears to rely on objective criteria.

[143] Art. 5(5) InfoSoc Directive.

[144] For a detailed argument on how an exception for user-generated content does not violate the second step of the three-step test, see JP Quintais, ‘Rethinking Normal Exploitation: Enabling Online Limitations in EU Copyright Law’ 41 (2017) 6 AMI: Tijdschrift voor Auteurs-, Media- & Informatierecht 197-205.

[145] R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), ‘Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition’ (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019.

[146] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 393, advancing a definition of non-commercial that “would prevent individual operators of P2P and other platforms that generate advertising or sponsorship or other financial consideration in connection with the usage of works from being exempt from liability as a result of the adoption of the statutory license.”

[147] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 288-295, 394.

[148] Id. at 394, noting that “[r]egardless, whatever indirect financial consideration end-users receive in grey-area cases, it is unlikely to cause substantial harm to right-holders, and should therefore be susceptible of compensation in the context of a statutory licence…”.

[149] Similarly, see R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019, 107, explaining that “a private user action may build upon previous acts of exploitation from third-parties that are covered by an exemption, however not through the exploitation of illegal file sharing.”

[150] For a detailed analysis, see JP Quintais, ‘Private Copying and Downloading from Unlawful Sources’ (2015) 46(1) IIC 66. For doubts as to the wisdom of imposing such a requirement for alternative compensation systems, see JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 390-391 and PB Hugenholtz and JP Quintais, ‘Towards a Universal Right of Remuneration: Legalizing the Non-commercial Online Use of Works’ in PB Hugenholtz (ed), Copyright Reconstructed: Rethinking Copyright’s Economic Rights in a Time of Highly Dynamic Technological Change (Kluwer Law International 2018) 266.

[151] R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019.

[152] M Senftleben and others, ‘The Recommendation on Measures to Safeguard Fundamental Rights and the Open Internet in the Framework of the EU Copyright Reform’ (2018) 40(3) EIPR 149.

[153] See Art. 11bis Berne Convention for the Protection of Literary and Artistic Works, July 14, 1967, 11850 U.N.T.S. 828, and Art. 12 Rome Convention.

[154] Arts 5, 6 and 8(2) Rental and Lending Rights Directive.

[155] See, on Art. 8(2) Rental Right Directive, Case C-245/00, Sena (6 February 2003) ECLI:EU:C:2003:68, [36]–[37]; Case C-192/04, Lagardère Active Broadcast (14 July 2005) ECLI:EU:C:2005:475, [50].

[156] See, respectively, Arts 5(2)(a), (b) and (e) InfoSoc Directive and Art. 6(5) Orphan Works Directive.

[157] For an overview of this case law, see e.g. A Dias Pereira, ‘Levies in EU Copyright Law: an Overview of the CJEU’s Judgments on the Fair Compensation of Private Copying and Reprography’ (2017) 12(7) J. Intell. Prop. L. & Practice 591.

[158] But see Art. 16 DSMD, recognising a claim of fair compensation for press publishers. In our view, this new claim, introduced in the context of the controversial new right for press publishers in Art. 15 DSMD, should not affect our proposal.

[159] Case C-572/13, Reprobel (12 November 2015) ECLI:EU:C:2015:750, [44]–[49]. See also European Copyright Society, ‘European Copyright Society Opinion on The Reference to the CJEU in Case C-572/13 Hewlett-Packard Belgium SPRL v. Reprobel SCRL’ (5 September 2015) https://europeancopyrightsociety.org/opinion-on-reprobel/ accessed 26 April 2019.

[160] See JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) and C Geiger ‘Copyright as an Access Right, Securing Cultural Participation Through the Protection of Creators’ Interests’ in R Giblin and K Weatherall (eds.), What if We Could Reimagine Copyright? (ANU Press 2017). 73.

[161] Case C-467/08, Padawan (21 October 2010), [40]–[42]; Case C-572/13, Reprobel (12 November 2015) ECLI:EU:C:2015:750, [36], [48]–[49].

[162] C Handke and others, ‘Going Means Trouble and Staying Makes It Double: the Value of Licensing Recorded Music Online’ (2016) 40 J. Cult. Econ. 227.

[163] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 293, drawing a parallel to the factors mentioned in Recital 35 of the InfoSoc Directive, which refer to the calculation of compensation for private copying.

[164] Id. 284-285, 293-294. See also J-J Vallbé and others, ‘Knocking on Heaven’s Door - User Preferences on Digital Cultural Distribution’ (14 July 2015) https://ssrn.com/abstract=2630519 accessed 26 April 2019 (forthcoming Internet Policy Review [2019]).

[165] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 394-397; J-J Vallbé and others, ‘Knocking on Heaven’s Door - User Preferences on Digital Cultural Distribution’ (14 July 2015) https://ssrn.com/abstract=2630519 accessed 26 April 2019 (forthcoming Internet Policy Review [2019]).

[166] R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019.

[167] See supra at 3.2.1.3.

[168] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 293.

[169] Id. at 381. On “appropriate reward” as an objective of EU copyright law and a notion central to the Court’s case law, see Recital 10 InfoSoc Directive and M Favale and others, ‘Is There an EU Copyright Jurisprudence? An Empirical Analysis of the Workings of the European Court of Justice’ (2016) 79 The Modern L. Rev. 31.

[170] See also R Hilty and A Bauer, ‘Use of Protected Content on Online Platforms’ in R Hilty and V Moscon (eds), Modernisation of the EU Copyright Rules Position Statement of the Max Planck Institute for Innovation and Competition (Max Planck Institute for Innovation & Competition Research Paper No. 17-12, 18 September 2017) https://ssrn.com/abstract=3036787 accessed 26 April 2019.

[171] An exclusion of this type is found for example in Art. 17(6) DSMD. According to this, certain online content-sharing service providers that have been operating for fewer than three years, have an annual turnover below EUR 10 million, and an average number of monthly unique visitors under 5 million are subject to “lighter” obligations under the liability regime of Art. 17(4) DSMD.

[172] Art. 17(6) DSMD mentions a threshold of EUR 10 million calculated in accordance with Commission Recommendation 2003/361/EC. However, a much lower threshold could be envisaged, if the objective were to foster the development of smaller platforms.

[173] In this system, CMOs “play a role in centralizing and monitoring information as well as in granting fair and efficient distribution (not neglecting the defence of creators’ interests against the copyright-based industries)”. See R Hilty and S Nérisson, ‘Collective Copyright Management and Digitization: The European Experience’ in R Towse and C Handke (eds.), Handbook on the Digital Creative Economy (Edward Elgar, 2013) 222-234, discussing the prospective role of CMOs in the digital environment. The authors suggest CMOs “could develop tools to enhance non-commercial uses of protected works thanks to micropayments and blanket licences”, thus enabling “compensation of creative people” and allowing “end-users sufficient access to copyrighted works”. In this respect, they add, “[e]xtending cases in which copyrights were to be statutorily collectively managed would help”.

[174] These would include data processing, information on uses, and distribution of rights revenue, to the benefit of second-level national CMOs (or subsidiaries) and other stakeholders.

[175] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 391-392 and 400-401. This leaves open the possibility, inter alia, that a CMO currently tasked with managing an existing exception, say private copying, would be tasked with also managing the new exception.

[176] This appears consistent with the objective expressed in Recital 11 InfoSoc Directive of providing adequate compensation to incentivise creation and safeguard the “independence and dignity of artistic creators and performers”. To ensure a fair distribution of revenue to right-holders, different contingent valuation methods could be combined, including techniques such as the measurement of use through electronic rights management information, sampling, and strictly anonymous monitoring with the assistance of content-sharing platforms. See JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 401, making similar points in the context of a broader proposal.

[177] See supra 3.2.1.2.

[178] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 401.

[179] M Senftleben and others, ‘The Recommendation on Measures to Safeguard Fundamental Rights and the Open Internet in the Framework of the EU Copyright Reform’ (2018) 40(3) EIPR 149.

[180] JP Quintais, Copyright in the Age of Online Access: Alternative Compensation Systems in EU Law (Kluwer Law International 2017) 245-325; JP Quintais, Rethinking Normal Exploitation: Enabling Online Limitations in EU Copyright Law 41 (2017) 6 AMI: Tijdschrift voor Auteurs-, Media- & Informatierecht 197-205. For a roadmap on a flexible interpretation of the three-step test, see C Geiger and others, ‘Declaration: A Balanced Interpretation of the Three-Step Test in Copyright Law’ (2010) 1 JIPITEC L. 119 para 1 and C Geiger and others, ‘The Three-Step Test Revisited: How to Use the Test’s Flexibility in National Copyright Law’ (2014) 29(3) Am. U. Int’l L. Rev. 581-626.

[181] PB Hugenholtz and others, ‘The Future of Levies in the Digital Environment: Final Report’ (Institute for Information Law, March 2003) http://dare.uva.nl/search?metis.record.id=223340 accessed 26 April 2019.

Fulltext

License

Any party may pass on this Work by electronic means and make it available for download under the terms and conditions of the Digital Peer Publishing License. The text of the license may be accessed and retrieved at http://www.dipp.nrw.de/lizenzen/dppl/dppl/DPPL_v2_en_06-2004.html.

JIPITEC – Journal of Intellectual Property, Information Technology and E-Commerce Law
Article search
Extended article search
Newsletter
Subscribe to our newsletter
Follow Us
twitter
 
Navigation