Copyright and Internet Service Provider “Liability”: The Emerging Realpolitik of Intermediary Obligations

Abstract

Online service providers are increasingly the focus of renewed efforts to enforce copyright online in the European Union (EU). Traditionally, these providers benefitted from safe-harbour immunity to the extent that their role in assisting online enforcement was relatively minimal. In light of recent proposals for reform, and the spread of the “blocking injunction” in the EU, this traditional position is coming under pressure. That pressure finds expression by way of a principle of accountability, leveraged by the entertainment industry and policymakers alike to carve out new obligations for these intermediaries, under the existing legal framework. This article focuses on the position of these intermediaries under Art. 16 of the European Charter of Fundamental Rights. Protecting the freedom to conduct a business, this provision has come to underpin a discourse of commercial fundamental rights, argued to have the capacity to curb the spread of obligations via accountability. This article takes a sceptical view, arguing that lessons from judicial approaches to the blocking injunction at the Member State level, indicate that the discourse of commercial fundamental rights may well overstate the strength of this rights-paradigm as a counter to obligations generally.

1 Introduction

When it comes to enforcing copyright in the online context, a changing Internet and changing attitudes that online service providers (OSPs)¹ should do more to assist with enforcement, go hand in hand. The more activity that has migrated online via their services, the more pressing a deceptively simple question for these providers has become: what can they do, or be forced to do, to manage those services in favour of enforcement? At the end of the last century this question may reasonably have been considered settled. All OSPs were, for all intents and purposes, non-liable for the actions of users on their network but were required to act in pre-defined circumstances.² The benefit was twofold: OSPs were positioned to leverage their nexus position to assist with enforcement but were not so encumbered as to stifle the innovative potential inherent in their services.³ The upshot for society has been profound, giving effect to everything from the “information age”, to the digital economy. But, as the need to regulate behaviour online – both civil and criminal – has only increased, this détente of lightly regulating these intermediaries is gradually becoming undone.

The question of whether (and how) OSPs should play a greater role has been led by the entertainment industry (the industry), either by way of political lobbying⁴ or litigation before the courts of the Member States. That debate centres on how far the subtle principle of accountability – inherent in the current legal framework – can displace its more pronounced sister-principle of non-liability with a view to broadening obligations. From policy perspectives, recent proposals for copyright reform in the EU provide something of an answer.⁵ Here, it is not hyperbole to suggest that proposed upload filters requiring intermediaries to effectively monitor their networks is an “earthquake” when it comes to OSP obligations.⁶ Not least, because EU law ostensibly prohibits such monitoring,⁷ but more broadly, because it points to a political will for OSPs to assume a greater role in enforcement modalities, despite the policy rationale of non-liability. With this lobbying success mirrored in national courts with the securing of orders against OSPs to implement the novel “blocking injunction”, there remains only one conclusion: obligations by way of accountability are on the march. Naturally, in a polycentric legal environment such as cyberspace,⁸ this does not occur in a vacuum. Indeed, while industry success points to the entrenchment of the property paradigm for copyright in this context,⁹ other stakeholder interests are also in play – acting as a counter-weight to industry ambitions. Until recently, the more pronounced have been the interests of Internet users in the shape of a counter-paradigm to that of copyright-as-property in the form of Internet user fundamental rights.

However, as the industry has pursued enforcement away from the user level, e.g. upload filters and the blocking injunction, it is the interests of OSPs that have taken centre stage. Here, the general position against a greater role in enforcement is grounded in both ideology and law. Ideological in that a shift away from non-liability – underpinned as it is with a cyber-libertarian rationale that neutrality for OSPs stimulates innovation – is a mistake. And, legal, in that the need to protect the commercial interests of these intermediaries is recognised by European primary and secondary law in this context. Article 16 then of the European Charter of Fundamental Rights (EUCFR) has, along with Art. 3 of the Enforcement Directive,¹⁰ come to underpin a discourse of commercial rights under the banner of the freedom to conduct a business.¹¹ While it is possible to argue that Art. 16 gives rise to a quasi-subjective right in this context – considerably inflating that discourse – this article takes a sceptical view. Focusing on selected judicial approaches and developments with the blocking injunction, it will be argued that, in practice, a paradigm of commercial fundamental rights may be less than robust in countering the march of obligations by way of accountability.

2 The Evolution in Intermediary Obligations: From “Liability” to Accountability

Pressure for OSPs to do more to assist with online enforcement, should in one sense be unsurprising; the development of the Internet from its early beginnings to where we are today could – of itself – be construed as a fait accompli. Nevertheless, the evolution in intermediary obligations is remarkable, not least because of the interplay between the principles of functional non-liability on the one hand and accountability on the other. Although both formed part of the early legal framework, non-liability emerged as something of a poster-child for best-practice regulatory approaches for OSPs in this context. However, although insulated from pecuniary liability, immunity was always conditional on OSPs acting in the interests of right holders. In short, non-liability may have been the poster-child, but the more subtle principle of accountability has underpinned the legal framework from the outset. As will be argued below, it is possible to view this conditional immunity as laying the foundation for the emergence and creep of obligations more generally. The starting point then is to consider the rationale and design of the safe-harbour regime.

3 The Starting Point: Liability and Immunity

That OSPs are non-liable for acts of online infringement is something of a misnomer. As such, under primary or secondary liability, an intermediary may well be found liable, provided the circumstances warrant it.¹² In practice however, most situations that might (in theory) give rise to liability – secondary in particular – are covered by conditional safe-harbours. Assuming that primary liability is the exception, and not the norm in the online context, e.g. that infringement is by the users of a service and not the provider, two approaches to the question of secondary liability are possible.¹³ The first – a positive approach – is to extend long-standing principles and doctrines of secondary liability that apply in the offline sphere to OSPs.¹⁴ In broad terms, this has met with little success, with the necessary evidentiary burden, proving – in the main – to be prohibitive for claimants. Typically, that burden requires a level of intent between the alleged secondary infringer and the actions of the primary infringer. This requires the claimant to prove a “common design”¹⁵ between both parties, to the extent that, not only did the OSP have knowledge of the wrong-doing, but actively conspired or induced the primary infringer.¹⁶ Although not impossible, e.g. where the intermediary has engaged in deliberate wrongdoing, the standard is typically not met in practice.¹⁷ The alternative then is to base the claim on the OSP’s knowledge of infringement occurring on its service. Here, the courts have ruled that general knowledge will not suffice.¹⁸ Rather, what is required is (i) contemporary knowledge of specific acts of infringement being brought to the attention of the intermediary by the right holder, and (ii) the continuation of service provision, after that knowledge has been established.¹⁹

In practice, although this may ground the claim, reasonableness enters the calculus to the extent that; if the OSP takes remedial action to mitigate, prevent or terminate the infringement, then it may well be deemed to have done all that was reasonable to avoid liability.²⁰ As Dinwoodie notes,²¹ judicial sensitivity in recognising these efforts is valuable because “the legitimacy of the behaviour of intermediaries occupies a spectrum that requires greater flexibility (and room for more subtle calibration) than formal secondary liability doctrine might seem to allow”.²² Significantly, this pragmatism also underpins the more dominant negative approach; the policy intervention in the shape of the safe-harbour regime itself, essentially obviating secondary liability.²³ Two mechanisms characterise that regime: (i) the mere conduit defence, and (ii) notice-and-takedown. Available to OSPs that act purely as transit intermediaries, e.g. access providers, the mere conduit defence recognises the passive nature of these services. In turn, so long as the provider does not initiate, select the receiver of, or modify a transmission, they will be deemed to be immune from liability²⁴ – albeit subject to injunctive relief.²⁵

By contrast, notice-and-takedown creates a non-passive dynamic for hosting OSPs, by way of positive threshold conditions for them to act to maintain immunity. These providers are therefore required to “expeditiously” remove infringing content, once made aware of its existence by notice from the right-holder.²⁶ A conceptual relationship can therefore be said to exist between the liability standards under the positive approach and the safe-harbours.²⁷ From a policy perspective, the safe-harbours then can be viewed as a form of business regulation, nudging OSPs towards mandatory conduct that would otherwise be required to avoid the imposition of secondary liability by the courts.²⁸ Non-liability may therefore be functional, but for present purposes, that immunity is contingent on its sister-principle within the legal framework: accountability. The significance of this is best illuminated by drawing out the policy rationale underpinning non-liability; the net social benefits of permitting OSPs to be as neutral as possible, outweighs the potential of positioning them as nodes of control capable of policing their own services.²⁹ Viewed under this lens, non-liability shaped by a policy position of neutrality is the default position, with accountability acting as insurance in the event that more would be required from these intermediaries in the future. Although it would not be unreasonable to suggest that accountability is entirely at odds with non-liability informed by neutrality, the reality is that obligations have been part of the equation here from the beginning. And, as circumstances have indeed changed, it is accountability that is coming to define the scope of OSP “liability” through the creep of obligations.

4 Accountability: The Creep of Obligations

Similar to the situation that prevails with notice-and-takedown, those obligations have arisen where the need dictates it, and the OSP has the capacity to act to provide either assistance or remedial relief. Accountability then is a tool of policy, capable of contouring obligations relative to developments in the online sphere. An exhaustive account of how this has unfolded is beyond the scope of this article. However, it is possible to bear out the point by demarcating the creep of obligations along the lines of (i) cooperative and (ii) quasi-regulatory modalities.

5 Information and the Cooperative Dynamic

In an era of “big data”, the prevailing concern surrounding OSPs and the information they transmit, is how much of that data is collected, stored and used to further their business models in a clandestine manner, detrimental to individual data privacy rights.³⁰ That OSPs have this capability is old news.³¹ But, when it comes to obligations, the most striking intervention remains the co-opting of OSPs by the State to retain customer traffic data for the purposes of counter-terrorism. With the inevitability of the Internet becoming a tool for terrorist propaganda, and an instrument of coordination for terrorist activity, a response by the State was perhaps unavoidable.³² That response took the shape of data retention legislation, requiring access providers to collect, store, and be in a position to make available the traffic data concerning all their subscribers.³³ Ultimately undermined as a result of data privacy rights of those subscribers,³⁴ data retention is nevertheless instructive for demonstrating the emergence of accountability in practice. Indeed, as Birnhack and Elkin-Koren note,³⁵ such developments signalled a new convergence between OSPs on the one hand, and third-party interests on the other – in this instance the State.

The significance lies not just in the scale of the obligation, but that the State intervened by co-opting the gatekeeping character of the OSP relative to its network, exploiting this to the extent of positioning access providers as nodes of control. Once positioned in this manner, access providers became, for all intent and purposes, agents for executing governmental power within a cooperative dynamic.³⁶ Cooperative mechanisms can therefore be defined as modalities under which an OSP is not required to take disruptive action against their network or users, but must nevertheless leverage their nexus position and disclose informational metrics concerning those users where called upon to do so. Not lost on the industry, once the paradigm of intermediaries as informational control nodes became established, a private sector convergence was pursued via disclosure of subscriber biographical data. Leaving aside the online context for a moment, where a third-party intermediary holds information that is necessary to pursue redress, the rule of thumb is that subject to an injunction, the intermediary may be ordered to disclose that information. Such relief exists in both common and civil law systems.³⁷ In the EU, this type of relief remains subject to data protection legislation. However, exceptions within the data protection framework, e.g. disclosure to fulfil a legal obligation and necessity to protect the rights of others, have ensured that neither data privacy nor data protection have been prohibitive to this aspect of industry strategy.³⁸

Over time, this relief has developed from an exceptional tool of justice to a fully-fledged remedy.³⁹ However, its utility for the industry remains the positioning of OSPs into this dynamic rather than the success of this particular modality per se. Indeed, owing to the high evidentiary threshold for successful litigation and the cost burden, the industry pursued the arguably lesser deterrent of letters claiming settlement.⁴⁰ Nevertheless, in overall terms, the pursuit of disclosure orders can be viewed as an indispensable tactic in furthering the creep of obligations generally. In this vein, such orders can be viewed as instrumental in demonstrating that the control node character of OSPs could be used, not just to the benefit of public, but also private stakeholders, when it comes to enforcement. Here, it is not unreasonable to argue that once demonstrated, this relief grounded a broader industry narrative that OSPs had to play a greater role in enforcement modalities if they were to be efficacious. Now entrenched, this conceptual view is something of a truism for policymakers and the judiciary alike, borne out by events after the industry’s strategy of pursuing such orders in the shape of the even more exceptional, quasi-regulatory interventions.

6 Control Nodes and Quasi-Regulatory Interventions

With the arrival of graduated response systems (GRS),⁴¹ OSPs faced a new prospect: forming an integral component of enforcement modalities that would take direct action against Internet users rather than merely collecting information and passing it on. Relatively short lived,⁴² the benefit to the industry was nevertheless – again – the positioning of intermediaries as key nodes, and the ability to leverage this through the framework for injunctive relief, once GRS effectively failed.⁴³ The mechanism to contour out accountability beyond notice and takedown, injunctive relief was always possible under the legal framework.⁴⁴ However, only in recent years has its true potential become apparent. To this end, as Hogan J of the Irish Court of Appeal has observed,⁴⁵ although the drafting of that framework left a lot to be desired, one thing is certain, the intent of European policymakers to vest national courts with the jurisdiction to grant orders for relief, that would otherwise be regulatory solutions. The significance of this cannot be overstated when it comes to industry strategy. Broadly speaking that strategy has always been one of disruptive regulation; interventions to test the margin of what may be acceptable and aligning future interventions with the lessons that have been learnt in a bid for more successful outcomes in the future.⁴⁶ With the emergence of Internet user rights as a counter-paradigm to online enforcement generally, policy support for measures at the user level are viewed as something of a political pariah. The injunctive relief framework allows for a way around this by pursuing intermediaries through the courts for quasi-regulatory solutions.

Using this framework then, the creep of obligations has shifted – in the main – away from the user level towards the non-user level⁴⁷ with the emergence of the blocking injunction.⁴⁸ The nuance of this relief will be considered in more detail below, but for present purposes, the significance lies in the fact that it is the judiciary, and not the legislature that is increasingly coming to flush out the scope of OSP obligations. This necessarily has consequences, particularly when it comes to demarcating the limits of what might be permissible under the legal framework. Not simply from the perspective of industry strategy, but significantly, in delineating a defensive line for OSPs against the imposition of regulatory burdens that may be deemed disproportionate. Indeed, this is especially so when it comes to positioning that line relative to notional commercial fundamental rights. In this respect, two approaches are possible. On the one hand, it is possible to argue that judicial reasoning here should be limited to the type of claim in question, e.g. the blocking injunction. If so, the value of that reasoning relative to broader questions surrounding the scope of OSP commercial rights would be limited. On the other hand, as Dinwoodie observes,⁴⁹ it is likely that the principles that have underpinned cooperative and quasi-regulatory solutions will give rise to, and shape, different measures in future technological and social conditions. With this outcome undoubtedly likely, it is inevitable that such principles will also radiate into the policy sphere. The question then is whether the “pragmatic hue”⁵⁰ that has come to characterise judicial reasoning here, can be reduced to a set of self-effacing norms that may indicate the scope of OSP commercial rights in practice. The starting point for this inquiry is Art. 16 of the EUCFR.

7 Grounding a Discourse of Commercial Rights: Article 16 EUCFR

Article 16 provides that the freedom to conduct a business, in accordance with Union law and national laws and practices, is recognised. That Art. 16 comes into play in this context is the result of CJEU jurisprudence in the opening gambit of the industry’s use of the injunctive relief framework in Scarlet.⁵¹ Here, the CJEU established the procedural approach that must be taken when considering the imposition of obligations under that framework. That approach recognises the constitutional equivalence of the right to protect intellectual property on the one hand, and the right of OSPs to conduct their business on the other. In turn, once constitutional parity is established, a balancing exercise must be employed, turning on the question of proportionality. The nuances of that balancing exercise, and the factors to be employed, will be considered in more detail below. But, in view of the fact Art. 16 grounds the discourse of commercial fundamental rights, it is necessary to consider not only its origins, but also its potential scope – namely, the extent to which it can be a vehicle for OSPs to resist the imposition of obligations.

8 The Origins of Article 16

Within the international legal order, the recognition of a freedom to conduct a business is unique.⁵² Indeed, rather than being drawn from the positive norms of that legal order, the freedom is drawn from the constitutional traditions of the Member States.⁵³ Although recognised by the CJEU since Nold,⁵⁴ the CJEU continues to offer little by way of “a full and useful” definition when it comes to the scope of the freedom.⁵⁵ However, its contour remains possible to demarcate along two aims: (i) to protect the economic initiative and activity of economic actors, and (ii) to ensure there are certain minimum conditions for economic activity on the internal market.⁵⁶ Those aims most closely align with the constitutional tradition of one Member State in particular: Germany. As Everson and Gonçalves note, in the post-war period, the German model of ordo-liberalism sought to protect an individual’s right to economic activity and set the conditions to promote this individual freedom by regulating the marketplace.⁵⁷ Thus, ordo-liberalism promotes the notion of an ordered economy, constituted and guaranteed by primary law and balanced against other social commitments.⁵⁸ Drawn from a cluster of rights, such as the right to property and the right to earn a livelihood, the freedom to conduct a business therefore entrenches protection for the economically driven individual.⁵⁹

Intrusions against this autonomy may occur, but these cannot be unwarranted, and will only occur where absolutely necessary. In contrast, the market is subject to relatively wide-ranging State intervention, such as competition policy to avoid concentrations of economic power that may restrict equality of opportunity.⁶⁰ In short; to promote fairness and balance in a social market economy. With the dual legal economic institutions of competition law and free movement, the EU Treaties and their underlying “Economic Constitution” can be said to be closely aligned to the fundamentals of ordo-liberalist thought.⁶¹ So much so, as Everson and Gonçalves argue,⁶² the recognition by the CJEU of a freedom to conduct a business in Nold, and its subsequent evolution to a general principle was a “chronicle foretold” when viewed in the context of the European project. Article 16 is therefore underpinned by a “desire to establish and maintain a political economy within the European Communities through the medium of law and the establishment of the Economic Constitution within the Rome Treaty”.⁶³ Be this as it may, its precise scope remains up for debate, with the dominant question being: is the freedom under Art. 16 to be construed as a guiding principle or a subjective right?

9 The Scope of Article 16

As such, it appears to be both. That Art. 16 may be construed as a principle and not a subjective right arises from a textual perspective of the provision itself, and CJEU jurisprudence. Taking the latter first, when it comes to that jurisprudence, the freedom may have risen to the status of a general principle, but in practice, it has had little effect in insulating economic actors from interventions that create regulatory burdens.⁶⁴ In turn, “whilst a commitment to free economic activity is recognised, the exercise of such rights must always be reconciled with other objectives – whether they are competing fundamental rights, or Union law more generally”.⁶⁵ The traditional approach of the CJEU has therefore been to recognise the freedom is in play, but qualify its scope by asserting limitations are possible, and invariably rule that the limitation in question is justified. Viewed through a lens of pragmatism, to do otherwise would be to allow economic actors a means – with a constitutional dynamic – to outflank the imposition of regulatory burdens that were in the interest of competing social market objectives. In practice, the recognition of the freedom then has been something of a damp squib. Nevertheless, the concern following the post-Lisbon EUCFR was that the inclusion of the freedom under Art. 16 elevated “a dangerously open-ended notion of business freedom to the normative level of fundamental rights”; with the potential to significantly broaden its scope.⁶⁶

However, following the CJEU ruling in Sky Österreich⁶⁷ such concerns may reasonably have been considered appeased. Here, the CJEU asserted that although the freedom was under the second title of the EUCFR dealing with individual freedoms, it not only differed from the wording of the other provisions of that title but it was more similar to certain provisions of the fourth title dealing with the principles of solidarity.⁶⁸ The freedom therefore can be said to have more in common with principles under the EUCFR than classical fundamental rights. In keeping with Art. 52(5) EUCFR, principles will not give rise to a direct cause of action against the Union’s institutions or the Member States. Article 16 therefore becomes relevant in the judicial setting where the implementing acts of either the institutions or the Member States are to be interpreted or are under review.⁶⁹ For good measure, the CJEU also affirmed not simply that Art. 16 should be construed as a principle, but informed by its own case law, that the freedom therein “may be subject to a broad range of interventions on the part of public authorities which may limit the exercise of economic activity in the public interest”.⁷⁰ As Grousset et al. argue,⁷¹ from a textual perspective, in limiting the scope of the provision, the CJEU appears to have placed emphasis on the wording of Art. 16 itself – that the freedom is recognised under the caveat “in accordance with Union law and national laws and practices”. In turn, not only is Art. 16 to be construed as a principle, but from a textual perspective, it is capable of being subjected to considerable restriction in contrast with other EUCR provisions.

Nevertheless, doubt remains that Art. 16 is to be construed in this manner – at least in all circumstances. Indeed, Grousset et al. argue⁷² that although the ruling in Sky Österreich suggests Art. 16, is to be construed as a principle, with its links to the right to property, freedom to contract and freedom to exercise economic activity, it nonetheless retains all the “prodromical signs” of a subjective right. In turn, even if its application in practice were limited, they argue Art. 16 should be justiciable in certain circumstances, e.g. in private disputes. Such an approach appears to immediately be at odds with the position in Sky Österreich, but the approach of the Court has been less than coherent. Indeed, notwithstanding the ruling, Art. 16 has been applied horizontally in private disputes. The most notable instances being the cases of Scarlet and Netlog,⁷³ where the Court considered the proposed monitoring and filtering systems in both cases to be a “serious infringement” of the OSP’s freedom to conduct their business. As Everson and Gonçalves suggest,⁷⁴ the only conclusion to be drawn is that Art. 16 can be said to have a dual character: a guiding principle one the one hand, and quasi-subjective right on the other. When it comes to the scope of Art. 16, this quasi-subjective right may therefore be argued to reflect an inflating of private autonomy⁷⁵ as it underpins the performative element of the freedom, e.g. as distinct from the right to establish commercial activity.⁷⁶ If that is the case, then Grousset et al. argue⁷⁷ it is possible private autonomy can be used to counter-balance “regulated autonomy” to the extent that Art. 16 may be invoked in horizontal situations as a guarantee of private autonomy. The effect of which would be to broaden its scope well beyond the ostensible limits established in Sky Österreich, allowing it to protect individual autonomy by limiting the full effectiveness of directives, as well as private regulation in general.⁷⁸ For present purposes, this means that Art. 16 retains the potential to undo the scope of accountability as a steering device for the imposition of OSP obligations – potentially giving significant teeth to the discourse of fundamental rights as it arises in relation to Art. 16. If a subjective right can be argued to be in play here, then a necessary question remains: to what extent can it blunt or curb the creep of those obligations?

10 Copyright, Accountability and the CJEU

With the purported quasi-subjective right to conduct a business drawn from the rulings of the CJEU in Scarlet and Netlog, the natural starting point for determining its scope should be the approach taken by the CJEU. If Art. 16 is to be a counter-balance against regulated autonomy, then it is within the balancing exercise employed by the court that its contour and substance should be illuminated. However, nuances at play in the particular context of online enforcement suggest that CJEU case law is not the vehicle to contour out the substance of the right argued to be in play. Nuance here broadly comes down to the CJEU’s penchant for vague and open-ended rulings, and the interplay between primary and secondary law in this context. Although Scarlet and Netlog provide a useful starting point for asserting that a quasi-subjective right might subsist, neither provide a firm basis as to the substantive content of that right. Much less, how it might go about challenging regulated autonomy in this context in any sort of coherent fashion. In contrast, case law from the Member States dealing with the EU legal framework as it applies to blocking injunctions provides relatively surer footing as to the scope of Art. 16 under this formulation. In order to understand why this is the case, it is first necessary to consider the various elements at play in proportionality balancing and how events have unfolded at the CJEU in the cases of Scarlet and Telekabel.⁷⁹

11 Proportionality Balancing

Where two or more rights come into conflict under the EUCFR, a balancing exercise is required by Art. 52, which provides that “any limitation on the exercise of the rights and freedoms recognised by this Charter must be provided for by law and respect the essence of those rights and freedoms. Subject to the principle of proportionality, limitations may be made only if they are necessary and genuinely meet objectives of general interest recognised by the Union or the need to protect the rights and freedoms of others”. Any quasi-subjective right under Art. 16 – as with all Charter rights – is therefore not absolute, and may be subject to limitation. In order to be justified, that limitation needs to be legitimate, suitable, necessary and proportionate.⁸⁰ In coming to this conclusion, the balancing exercise is comprised of at least four steps: the first being to consider the legitimacy of the basis in a particular context for limiting one right over another; the second, to judge the suitability of the measure that will give effect to the limitation in question; the third to judge the necessity of the measure; and the fourth the actual balancing itself.⁸¹ In theory, each stage has the capacity to blunt the imposition of the proposed limitation; a measure that may be said to be either unsuitable or illegitimate or ineffective would preclude the need to go to the balancing stage.⁸² In practice however, the first three stages are easily discharged, this coming down to the constitutional parity that is said to subsist (at least in the CJEU)⁸³ between the various EUCFR rights and the Court’s general approach to resolving conflicts exclusively as an issue of fundamental rights.⁸⁴

Legitimacy is therefore satisfied on asking if the proposed measure is for a proper purpose, suitability on whether the measure is capable of at least partially fulfilling the objective, and necessity by asking if there are no other less impacting measures capable of fulfilling the objective.⁸⁵ With copyright functional under a banner of constitutionalisation, a proper purpose is easily satisfied, with a low threshold applied to suitability and necessity.⁸⁶ This can be criticised for allowing measures that are ostensibly ineffective or indeed unnecessary to reach the interpretative balancing stage.⁸⁷ But, a certain degree of pragmatism is at play in allowing a consideration of all measures rather than creating chill effects. Nevertheless, as will be seen below, once in the fourth stage, and interpretative balancing commences, the property dynamic at play continues to have effects. Here, the zone of proportionality or leeway available to a decision-maker is characterised by four possible discretionary approaches when it comes to the intensity of review to be applied to a particular measure. In broad terms, the balancing stage is concerned with an assessment as to “whether the attainment of the legitimate aim balances the limitation of interests necessarily caused by the act in question”.⁸⁸ This essentially means that a measure may be legitimate, suitable and necessary, but can still be disproportionate as to its effect on a particular right.

In weighing whether or not this is the case, proportionality as a doctrine is not rigid. Rather, it is a “flexi-principle”⁸⁹ by virtue of the variable intensity of review that can be employed by the decision-maker in weighing how serious a limitation might be acceptable relative to the purported gain.⁹⁰ Key to this flexible principle then are the four discretionary approaches that can be employed in the calculus of achieving balance: structural, policy-choice, cultural, and empirical discretion.⁹¹ The extent to which any one discretion dominates comes down in large part to the nebulous concepts of judicial deference and restraint.⁹² In the context of review, this may be said to be between the reviewing court and the decision-making body, but the analogy also holds when it comes to the interplay between the CJEU, the EU institutions, and the Member States. As Rivers argues,⁹³ in order to understand the variables at play when it comes to the intensity of review, a doctrine of deference (and restraint) must be recognised as a principle, at play in its own right, alongside proportionality. How deference unfolds should be contingent on the seriousness of the proposed limitation; the greater the burden, the less deference should be in play.⁹⁴ Institutional competence and margins of decision-maker discretion should, therefore, not be prohibitive to deference being reduced. If the limitation over-steps the mark, then the onus is on the judiciary to assert their authority within the institutional order to protect the right at stake. In turn, deference is a question of degree as to the extent of respecting the measure under review, and restraint is not setting aside measures or alternatives that are ultimately proportionate.⁹⁵

12 Scarlet: Ordo-Liberalism or Aberration?

In Scarlet, the view of the Court was that to order the Internet service provider (ISP) to implement a filtering and blocking system for online infringement on its network would be a “serious infringement” of the ISP’s freedom to conduct a business.⁹⁶ The basis of this coming down to the Court’s proportionality balancing placing an emphasis on the costly nature of the modality, and that the ISP would have to shoulder the costs involved, entirely “at its own expense”.⁹⁷ In broad terms the ruling can be viewed as a departure from the deference the CJEU has typically shown to the policies supporting the enforcement of copyright generally. Indeed, although the Court has been unequivocal that intellectual property rights are not absolute, it nevertheless has been consistent that such rights deserve a “high level of protection”.⁹⁸ Moreover, when it comes to the regulatory burden that may be imposed as a result of pursuing Community objectives for certain economic stakeholders, the Court has ruled that “even substantial negative consequences” may be justified relative to the goal of the regulatory burden.⁹⁹Scarlet then is viewed as a watershed moment in which the Court refused to expand the boundary of online enforcement to include active monitoring and blocking of real-time communications. Arguably more concerned with establishing a ceiling point to protect Internet users,¹⁰⁰ the import for ISPs appeared to be the same in insulating them against this type of relief. However, in the context of Art. 16, questions remain as to the scope of the ruling concerning obligations generally. Indeed, as AG Villalón has observed,¹⁰¹ the freedom to conduct a business is not concerned with protecting profit, but rather the economic actor’s “economic initiative and the ability to participate in a market”. In the absence of reasoning by the CJEU to the contrary, it is not altogether clear that the cost burden contemplated by the relief in Scarlet, would have had the effect of negatively impacting the ISP’s ability to participate in the market. That the CJEU refused to grant the relief because of the scale of the system required relative to the OSPs’ freedom to conduct a business therefore remains questionable.

Rather, as Everson and Gonçalves argue,¹⁰² they may have been able to continue trading, but the cost burden could be construed as simply unfair. To this end, they argue that – framed by a concern for fairness – the CJEU ruling can be construed as a principled application of Art. 16 in shielding the economic actor in question from “unfair business costs”.¹⁰³ An ordo-liberal rather than neo-liberal approach then, designed to prevent the ISP from being put at a competitive disadvantage.¹⁰⁴ Taking this interpretation to its logical conclusion however, the threshold for discharging a limitation as a “serious infringement” based on the variable notion of fairness and relative competitive disadvantage could, ultimately, be remarkably low. If so, the regulatory framework, as it applies to intellectual property and OSP obligations, could fast become undone. Deference to institutional competence,¹⁰⁵ along with the nuance at play in Scarlet suggest this was not the intention of the Court. Nuance here being the broader effect of the modality on Internet user rights,¹⁰⁶ and the possibility that the light treatment of Art. 16 was a means to shore up the rejection of a system that would filter and block real-time communications. In turn, the better starting point for construing the scope of Art. 16 may be the CJEU’s ruling in Telekabel¹⁰⁷ rather than Scarlet.

13 Telekabel: A Return to Pragmatism?

At issue here was whether the ISP in question could be required to bear the cost of the blocking injunction being pursued against it. If the Court was, in fact, pursuing a principled application of Art. 16 based on unfair business costs, it would not be unreasonable to expect the Court to have taken the opportunity to anchor in this purported general approach to Art. 16 in Telekabel. However, recognising that the injunction would (i) restrict the ISP’s free use of its resources, (ii) represent a “significant cost”, (iii) considerably impact on the organisation, and (iv) require difficult and complex technical solutions,¹⁰⁸ the Court nevertheless ruled the relief could be granted. The basis was that “such an injunction does not infringe the very substance of the freedom of an Internet service provider such as that at issue in the main proceedings to conduct a business”.¹⁰⁹ This was informed by two factors. In the first instance, it was for the ISP to determine the specific measures to be taken to achieve the result sought, adapted to the resources and abilities available to it, and compatible with other obligations the ISP will encounter in the exercise of its activity.¹¹⁰ In in the second, liability could be avoided by proving that all reasonable measures relative to the ISPs resources were undertaken.¹¹¹ The effect of which is that the ISP will not be required to make unbearable sacrifices, which as the Court notes “seems justified in particular in light of the fact that he is not the author of the infringement of the fundamental right of intellectual property which has led to the adoption of the injunction”.¹¹² The measures in question therefore need to be all that could reasonably have been expected from the ISP relative to its resources.

The Court’s position in Telekabel then sits in stark contrast to the relatively vague treatment of Art. 16 in Scarlet. Indeed, in recognising that the regulatory burden on the ISP may be a significant limitation of its freedom to conduct a business, that limitation could still be justified where proportionate, even where costs and complexity were significant. Aligned more closely with the original contour of Art. 16, Telekabel suggests the discourse of fundamental rights surrounding OSP obligations has limits. The question is, how much? In one sense, it is possible to argue that there are no limits per se, by virtue of the CJEU’s institutional function, and its unfettered discretion to weigh the proportionality of measures that might arise under accountability via Art. 52 EUCFR.¹¹³ But arguably, this generally unfettered discretion is curbed to a certain extent by way of the Enforcement Directive.¹¹⁴ In terms of institutional competence, the Enforcement Directive is clear as to the will of the legislature that relief must be available to right-holders to terminate or prevent infringements.¹¹⁵ Entrenching not only the principle of accountability within the overall policy framework, but also the over-arch of a proprietary dynamic when it comes to those obligations. Key is Art. 3, which provides that any measures or remedies adopted must be: (i) fair and equitable, (ii) not unnecessarily complicated or costly, and (iii) must be effective, (iv) proportionate, and (v) dissuasive.

It is possible to argue that these threshold conditions are mere specifications of the broader principles of reasonableness and proportionality underpinning Art. 52(1). If so, CJEU discretion would remain unfettered, with the scope to inflate Art. 16 remaining intact.¹¹⁶ In turn, the real strength of Art. 16 would lie in this unfettered discretion and the CJEU’s capacity to invoke the provision to shield OSPs against the creep of obligations.¹¹⁷ In practice, however, it is possible to argue that secondary legislation here dominates to the extent that the Enforcement Directive shapes judicial calculus to the degree that OSP obligations are the will of the legislature, that will entrenches accountability and the need to protect intellectual property, and the threshold conditions for those obligations are laid down in Art. 3. In practice then, Art. 3 has come to create something of a sui generis situation in which secondary legislation dominates, creating a broad basis for judicial deference to the imposition of obligations, and the threshold conditions for establishing them. While this may be criticised as suggesting the tail wags the dog,¹¹⁸ the reality is that the scope of Art. 3 covers the contextual basis of the balancing exercise in this context. Indeed, the experience at the national level to-date suggests that the criteria have proven relatively sound in flushing out the contour of the fundamental rights paradigm as it applies to OSPs in this context. In contrast, although CJEU rulings such as Telekabel can be said to establish the outer contour of Art. 16 at play here, e.g. that a limitation must go to the substance of the freedom, little is offered by way of pinning down that substance. In the sui generis context of online intellectual property enforcement, it is arguably the approach taken by the Member States to Art. 3 and not that of the CJEU that therefore offers a more accurate barometer as to the scope of Art. 16 when it comes to balancing obligations on the one hand, and commercial fundamental rights on the other.

14 The National Courts: Idealism vs. Pragmatism

At the national level, the industry has had marked success in securing injunctive relief against OSPs.¹¹⁹ Rejecting that the relief touches on the upper ceiling point, e.g. fundamental rights such as freedom of expression, national courts have – in the main – had little difficulty in granting relief. In line with how proportionality balancing typically proceeds, legitimacy, suitability and necessity are quickly discharged, with two criteria under Art. 3 coming to dominate the judicial calculus – effectiveness on the one hand, and cost (and complication) on the other. Taken together, both criteria go to the core of answering whether a limitation is an interference with Art. 16 that goes to the substance of the freedom. As such, if a measure is unnecessarily costly and complicated, or ineffective, forcing the OSP to implement it would be unreasonable because the result would be unjustified relative to Art. 16. The question is, to what degree this holds true in both circumstances.

15 The Question of Efficacy

In requiring that measures be “effective”, Art. 3 of the Enforcement Directive not only equips the discourse of commercial fundamental rights with potential teeth, but the judiciary with a nuclear option to blunt OSP obligations. The logic here is relatively simple; proving the effectiveness of a measure when it comes to enforcement can – at best – be difficult, not least because of the relative anonymity of users. A degree of uncertainty as to efficacy therefore afflicts all modalities that are aimed at enforcement. Where uncertainty as to the outcome of a particular measure is present, then the “laws of balancing” suggest that when it comes to proportionality, the greater the uncertainty as to the outcome of limiting a particular right, then the greater the need for empirical data to support the move.¹²⁰ This uncertainty is therefore argued to be a sufficient justification for adjusting the variable intensity of review in requiring any measure to be demonstrably capable of achieving its objective. The import of which is that the scope for judges employing policy-based discretion, implicitly shaped by deference to the will of the legislature, should be curbed, in favour of empirical-based reasoning.¹²¹ If so, then the industry should have a serious problem. In short, if effectiveness was construed in this strict manner, the threshold for imposing obligations – either by litigation or as a matter of policy – would be relatively high, if not prohibitive in most circumstances. In an ideal world that is – the reality has been far less kind to this approach. Indeed, although the blocking injunction appeared to offer fertile terrain as a tailor-made factual matrix in which to push for a high threshold for solid evidence-based law-making, this has not been borne out.

16 The Ziggo v. Brein Saga

In an early test as to the role effectiveness under Art. 3 could play in blunting the imposition of obligations, the Dutch Court of The Hague threw down a gauntlet to the industry – and European policymakers more broadly – overturning a block granted by the lower court on the basis of evidence bearing out ostensible ineffectiveness.¹²² The position of the lower court¹²³ was that the European legal framework permitted relief that was preventative, and not just targeted against existing or known infringements.¹²⁴ That the site in question, the “PirateBay”, was made up predominantly of illegal content and that the balance of proportionality lay with making the order. In this regard, it was ruled that the order could indeed prevent infringement, and was necessary in a democratic society to safeguard the rights of copyright holders. The order imposed was against two of the Netherland’s largest ISPs; Ziggo and XS4ALL to block a total of 24 websites. Immediately, the order was appealed to the Court of the Hague¹²⁵ on the basis that the lower court had erred in its reasoning when it came to interpretative balancing; the order was therefore argued to be disproportionate vis-à-vis Arts. 3 and 16 EUCFR. In weighing this argument, the Appeal Court of The Hague considered that Arts. 17(2) and 16 of the EUCFR should be read in light of Art. 52(1), with the question of proportionality turning on the criteria under Art 3 of the Enforcement Directive.¹²⁶ On this basis, in evaluating the order as to its effectiveness, the court considered that “effectiveness” would have to be judged on whether or not (i) the persons infringing copyright did so less, so that the number committed actually falls, e.g. in overall terms, or (ii) the impact of the infringements is reduced because of a diminution in the public that can actually see such illegal content.¹²⁷

Accepting as a matter of fact that the number of visits to the PirateBay were reduced to a “significant extent”,¹²⁸ the court nonetheless held the blockade was ineffective as Internet users can “dodge” the blockade¹²⁹ by using proxy addresses¹³⁰ and virtual protected networks¹³¹ (VPNs) or they may access materials BREIN sought to protect through alternative websites. In turn, given that the block was to protect such works, it was ineffective in this “contemplated objective” by way of circumvention measures. In this regard, the court relied on a report from researchers of the University of Amsterdam¹³² measuring BitTorrent downloads during the period of the block, finding that BitTorrent traffic remained unchanged on the ISP’s networks, notwithstanding direct access to the PirateBay had diminished. To this end, that study concluded that the block in question could only affect 27–28% of consumers pirating content. Moreover, of this number, 70–72% were found to be unresponsive to the block in finding alternative means to access the same, or different sites. In turn, the block only affected 4–6% of the network’s users.¹³³ In sum, the block had little effect on the overall levels of piracy occurring on the networks. Moreover, the court held that while the ISPs did not challenge BREIN’s assertion that the block had little effect in terms of cost, it remained an infringement of their freedom to conduct a business at their own discretion under Art. 16 of the Charter.¹³⁴ Given the ineffectiveness of the block was accepted by the court, it further held imposing an ineffective enforcement modality on the ISP compounded the lack of proportionality or a justification for interfering with this discretion.¹³⁵

17 The General Approach

In light of the Court of The Hague’s ruling it is clear that where so inclined, effectiveness can indeed be leveraged by the judiciary to blunt obligations – a powerful tool then for OSPs and those who wish to challenge enhanced obligations generally. The only problem being the ruling (now overturned)¹³⁶ is in the minority. Rather, the general approach – adopted by national jurisdictions¹³⁷ and the CJEU¹³⁸ – has been to position the threshold for effectiveness relatively low.¹³⁹ In this regard, arguments that the blocks do little to reduce overall levels of infringement, e.g. in aggregate relative to piracy occurring on the network as a whole, have been rejected.¹⁴⁰ That the ineffectiveness of such blocks can be borne out by easy circumvention has also been rejected. According to the CJEU,¹⁴¹ an absolute block is not feasibly practicable; the block may be capable of circumvention, but it nevertheless creates a barrier or a hurdle that must be overcome. Just because users can overcome this barrier does not mean that the measure is not sufficiently effective to ensure genuine protection of the fundamental right at issue.¹⁴² Therefore, so long as the block makes the accessing of piracy-facilitating websites difficult and can “seriously discourage” users from accessing the material therein, effectiveness is satisfied.¹⁴³ As noted, this approach has now been followed throughout the EU, including by the Court of The Hague,¹⁴⁴ where under the direction of the Dutch Supreme Court,¹⁴⁵ it re-imposed the block on the basis that, inter alia, its reasoning as to effectiveness had erred; the general approach was subsequently applied.¹⁴⁶

18 A Hybrid Discretion

When it comes to the discretion employed by the judiciary in coming to this conclusion, something of a quasi-hybrid position between empirical and policy-based discretion has emerged. In this sense, the judiciary have made use of metrics to indicate that blocking does have an effect, with a consensus that direct access to a blocked site can be achieved somewhere in the region of 70%.¹⁴⁷ Less conclusive has been the effect of the block on user behaviour when it comes to circumvention. Here, the metrics are open to relative interpretation to the extent that some studies demonstrate an effect,¹⁴⁸ while others suggest the effect of the blocks when it comes to circumvention is minimal.¹⁴⁹ That neither is dispositive has created room for the judiciary to cast a veneer of empiricism over – what otherwise appears to be – policy-based discretion. In turn, the reliance on studies more favourable to the suggestion that the blocks have an effect, allows for the proposed block to be weighed against empirical metrics, suggesting that the variable intensity of review has been adjusted to account for the relative uncertainty inherent in this context. In one sense, there is no reason not to take these rulings at face value; in focusing on the effect of the individual block rather than the impact of that block on overall levels of piracy, the argument that the block is effective, albeit capable of circumvention, is borne out. However, a significant degree of nuance is at play. While it is true that not all users will circumvent, studies such as that of the University of Amsterdam cast a pall over the consensus that effectiveness is calibrated sufficiently high in order to justify the limitation to OSPs’ commercial fundamental rights vis-à-vis Art. 16.

But, this is precisely the point. With metrics unfavourable to the imposition of a block rejected in favour of metrics and reasoning that allows for their imposition, it is arguable a quasi-hybrid discretion has been employed, comprised of empiricism on the one hand, and policy on the other. With the threshold for effectiveness in this context set relatively low, two points emerge: (i) the judiciary can be said to have aligned themselves to the deterrence rationale that underpins industry strategy when it comes to online “enforcement”, and (ii) this can be argued to have consequences beyond the blocking injunction. Underpinning this subtle alignment then is a pragmatism, informed by the constitutionalising of intellectual property generally, but more pointedly, deference. In this regard, should effectiveness be inflated to require a firm evidentiary basis beyond a low threshold, its capacity to blunt not only obligations but policy writ large becomes a real prospect. Depending on your ideological leaning, this may be the whole point. However, in light of the difficulties in bearing out the effectiveness of modalities with any degree of certainty one way or the other, it is perhaps sanguine to expect effectiveness to be set at anything beyond a relatively low threshold. To do otherwise would be counter-intuitive to policy, subject to instances where the modality in question is blatantly ill-equipped for its ends.¹⁵⁰ The signal then in the context of measures to enforce copyright online? As a criterion in the balancing between limitations on the one hand, and Art. 16 on the other, effectiveness will not be inflated to the degree that it will sink policy to the extent of frustrating modalities that have even a de minimis effect.¹⁵¹ If so, then this leaves cost (and to a lesser extent complication)¹⁵² as the only real hurdle within the discourse of commercial fundamental rights to be overcome when it comes to secondary liability and OSP obligations.

19 Cost and Complication

If cost (and complication) emerge as the only real factors that have the capacity to shape obligations relative to Art. 16, this brings us back to where the line should be drawn when it comes to the “substance” of that freedom. On this issue, the legal framework is open to interpretation one way or the other. In this regard, experience in the UK, in particular, has shown that the framework, and pronouncements of the CJEU can be read as supporting the view that the intermediary should bear the costs of implementing a particular enforcement modality. But, recent developments have pointed to the fact that more nuance is at play here, namely that EU law may only support this view if national law does not prohibit the imposition of costs. Where the balance should lie absent harmonisation, remains open to debate. But, as will be argued below, we may already have a signal from the CJEU as to the preferred approach, following its ruling in Telekabel.

20 The EU Legal Framework: Implying the Imposition of Costs?

The initial position in the UK was that the rightholder (the industry) should bear the costs of an unopposed application, while the intermediary should bear the cost of implementing the order, e.g. the running costs of the modality itself.¹⁵³ The logic here is relatively simple. As Arnold J has ruled,¹⁵⁴ ISPs are a commercial enterprise that make a profit from the provision of services, which are in turn used by operators, e.g. websites such as the PirateBay, and users to infringe copyright. If so, although there is no question of wrongdoing on the part of the OSP, costs can nevertheless be construed as a cost of carrying on business in this sector,¹⁵⁵ and accordingly, imposed. Arnold J’s legal basis for this comes down, as the latter suggests, to commercial equity,¹⁵⁶ but also from the EU legal framework and rulings of the CJEU. When it comes to the legal framework, according to Arnold J, it is implicit in recital 59 of the InfoSoc Directive that the European legislature has chosen to impose the costs on the intermediary.¹⁵⁷ That recital provides inter alia that a rightholder should have the possibility of applying for an injunction against an intermediary on the basis that in many cases, such intermediaries are best placed to bring such infringing activities to an end.

Moreover, in Cartier¹⁵⁸ he relied on the CJEU’s reasoning in Telekabel,¹⁵⁹ where it ruled that

an injunction such as that at issue in the main proceedings constrains its addressee in a manner which restricts the free use of the resources at his disposal because it obliges him to take measures which may represent a significant cost for him, have a considerable impact on the organisation of his activities or require difficult and complex technical solutions.¹⁶⁰

The inference being that the cost burden could be imposed on an intermediary, with proportionality naturally being a matter of degree. Moreover, for the purposes of interpreting the scope of European law, this could be construed as a principle to be applied in cases of injunctive relief concerning OSPs generally. On appeal, the UK Court of Appeal agreed with Arnold J,¹⁶¹ going so far as to conclude that the price of immunity for intermediaries was the burden of costs for enforcement modalities.¹⁶² The costs of compliance were therefore a quid pro quo, and could be inferred not only from recital 59 InfoSoc, but also recital 23 of the Enforcement Directive, both of which ground the legal right to apply for the injunction. The UK Supreme Court,¹⁶³ however, has taken a different view; namely that inferring from the EU legal framework that the EU has somehow harmonised, or mandated, that intermediaries should bear the cost burden for enforcement modalities goes too far.

21 Costs: A Matter for National Law

To this end, the directives are silent on the specific issue of where the costs should fall when it comes to such modalities.¹⁶⁴ On the contrary, recitals 59 and 23 explicitly state that the conditions and modalities/procedures relating to such injunctions should be left to the national law of the Member States. Relying on this, the Supreme Court ruled that imposing the cost burden on the ISPs is contrary to English law.¹⁶⁵ Here, English law is governed by rules that have developed surrounding its Norwich Pharmacal relief for orders for information disclosure, where the applicant, and not the innocent intermediary bears the costs for the relief granted.¹⁶⁶ With the EU legal framework silent as to costs, and deferring to national law, whether this would be the case in other Member States comes down to national laws on a case-by-case basis. Dealing with the CJEU rulings relied on by Arnold J and the Court of Appeal, the Supreme Court took the view that those rulings were therefore inapplicable in this situation.¹⁶⁷ When it came to Telekabel, the Supreme Court disagreed that the CJEU had laid down any principle concerning the imposition of costs that might be applied in line with the EU legal framework. Rather, its determination that costs could be applied, was specific to the preliminary in question, which was dealing with Austrian law, which unlike English law, does make provision for the intermediary to cover the costs of implementation.¹⁶⁸ As it applies to costs, that framework can therefore be construed as envisaging disparity among the Member States as to whether or not costs will be imposed as a matter of national law.¹⁶⁹

22 Cost, Complication and Article 16

When it comes to the discourse of commercial fundamental rights, it is apparent then, that unlike effectiveness, these dual criteria may well ground that discourse on more concrete terms. In light of recitals 59 and 23 of the InfoSoc and Enforcement Directives respectively, the degree to which this holds true will depend on the national law of the Member State. However, although the UK Supreme Court distinguished Telekabel as interpreting Austrian national law, the approach taken there by the CJEU has a broader import, both for national approaches, and possible future harmonisation. Key is how the Court dealt with the issue of Art. 16 and the substance of the freedom. While the ruling has been critiqued for leaving it to a private entity to determine the scope of measures that touch on fundamental rights, the fact remains that in doing so, the CJEU was attempting to achieve balance. Costs could be imposed, but the shape of the modality was to be relative to the overall resource base of the undertaking. If an ordo-liberal leaning is arguably at play in the CJEU’s rulings, then it is more appropriate to position it here rather than, as Everson and Gonçalves argue, with Scarlet. If that is the case, then this opens up scope for manoeuvre when it comes to Member States that allow for the imposition of costs, and any potential future harmonisation – namely, when it comes to rationalising fairness and curbing cost and complication as the dominant elements in underpinning the discourse of commercial fundamental rights. Key then is how fairness is to be rationalised in this context.

This brings us back to the position of the UK courts. Although the Supreme Court was merely following domestic rules surrounding costs, the broader import of this for online copyright enforcement and OSP obligations generally, has not been missed. As Husovec argues,¹⁷⁰ in having the industry pay the costs of implementing and maintaining a modality, it may well prove more efficient in forcing the industry to conduct a cost–benefit analysis relative to effectiveness prior to making a claim. If ineffective, and accordingly incapable of justification relative to cost, then ineffective measures will likely not be pursued. Indeed, as Husovec observes

costs are inextricably linked to the very jurisdiction to grant the measures. The allocation creates an impression of the conceptual context of the jurisdiction, and its daily use is influenced by it. There, to get the injunctions against innocent third parties right, we need to get the costs right too. Other European countries might take Cartier as an opportunity to reflect on their own systems, which are often the result of formalist inertia, rather than intentional design.¹⁷¹

The logic to this argument is sound but it remains an open question as to whether Member States that do allow for cost, much less the CJEU when it comes to a potential harmonisation, will be influenced, or indeed follow, the UK approach.

Notwithstanding the position of the Supreme Court that Telekabel laid down no general principle as to the allocation of costs, where costs are permitted by a Member State, then the CJEU’s position in Telekabel is likely to dominate vis-à-vis fairness. Similarly, there is no reason to consider that the CJEU’s approach in a potential harmonisation move would be any different. In the first instance, with Art. 3 of the Enforcement Directive envisaging cost and complication that is not unnecessary, it requires a short deductive leap to find a potential basis for that harmonisation. And, in the second, it is dubious that, in the broader context of the EU’s goals surrounding intellectual property policy generally, along with the constitutionalising of the right, that the CJEU would take a narrow view of the public interest here.¹⁷² In this regard, for the Supreme Court, the benefit of enforcing the right via the blocking order was said to be squarely for the benefit of the industry, limiting the public interest benefit to the industry and therefore underpinning the efficiency argument above. But, the contrary view is also possible, that the enforcement of intellectual property advances both the public and private interest, creating space for a consideration of costs in line with Telekabel.

If that is the case, then it remains an open question as to what might satisfy the signal from the CJEU when it comes to fairness. Pragmatism appears to be as good a place as any to start, and this brings us back to the reasoning that fairness here can be supported by the customer base of the OSP or through cost-sharing. In the first instance, as Arnold J has observed, it is possible to simply construe the cost burden as a cost of doing business in this sector.¹⁷³ In a zero-sum analysis, the OSP benefits from the use of its service by its users to engage in piracy – as it would where those services were used for more legitimate ends. Underpinning this is the argument expressed by the UK Court of Appeal and considered in the Irish High Court as having “considerable strength” that the price for immunity is the burden of costs as implied by the EU legal framework.¹⁷⁴ In short, the customer base of the OSP affords it the opportunity to amortise the costs over time. As such, this reality supports a counter-view as to efficiency as outlined above. As Arnold J has observed “requiring intermediaries to pay the incremental costs of implementation gives them an incentive to be efficient, to innovate, and to co-operate with rightsholders in order to keep costs down”.¹⁷⁵ That said, if fairness becomes the driving force of the analysis here, then the conclusion that the industry benefits from enforcement – regardless of the intersecting public – may alter the calculus.

This was the case in the Irish High Court where the costs to be imposed on the ISP, UPC Ltd, was of an order of magnitude higher than the implementing costs of a blocking order. Here, in order to achieve fairness, Cregan J ruled (as Arnold J had indicated he would if the factual matrix before him warranted it)¹⁷⁶ that cost-sharing between the ISP and the industry was appropriate, e.g. to a ratio of 80:20 in favour of the industry.¹⁷⁷ In the context then of costs, it is apparent that, for now, whether the cost burden will be imposed will depend on the Member State in question. Where, like in the UK, no costs are to be imposed, then this may have the efficiency effect that Husovec argues should be of broader application within the EU. On the other hand, where Member States do allow for costs, it is likely that in order to be fair, a certain deference will have to be paid to the signal by the CJEU in Telekabel. How that unfolds will depend on what can be justified as fair in the particular circumstance. But, it is important to recall that if Art. 16 comes into play, then the “substance of the freedom” comes down, not to profit per se, but to the ability to participate in the market. As was the case in Telekabel, there appears to be significant scope for the OSP to take on obligations, without this threshold being crossed. And, as evident from the line of reasoning that sees cost in this sense as a cost of doing business, options remain for achieving fairness in this regard. In overall terms then, it remains possible that the issue of costs can anchor the discourse of fundamental commercial rights to a factor that can shift the calculus in favour of the OSP in a given circumstance. But, in view of the CJEU’s reasoning in Telekabel, and the variable outcomes permissible in the various Member States, it remains dubious that the criteria will have any real impact in undoing the creep of obligations.

23 Conclusion

As noted from the outset, it is possible to dismiss the notion that developments surrounding the blocking injunction can indicate the broader scope of the discourse surrounding commercial fundamental rights. This would be a mistake. Not least because at play within this reasoning is an emerging realpolitik when it comes to enforcing copyright online. Accountability sets the stage for an over-arching entrenchment of the need to protect intellectual property, not absolute, but deserving of a high level of protection. Once that stage is set, any proportionality analysis – either judicial or policy-based – is framed by an implicit hierarchy when it comes to weighing stakeholder interests. At the apex of this hierarchy is the Community interest in securing the protection of intellectual property relative to broader socio-economic and cultural goals. As seen, when it comes to proportionality balancing, the initial criteria of legitimacy, necessity and suitability are easily satisfied, not least because of this over-arching goal of the Community – leaving us with efficacy and cost as the dominant criteria to be discharged. The realpolitik at play here is most pronounced relative to these criteria. Indeed, efficacy alone, from its potential to how it has functioned in practice, is indicative that the only real concern comes down to cost. How much of a bearing that will ultimately have on blunting the creep of obligations comes down to the particular modality in question. But, if fairness (with an ordo-liberal slant) comes to frame the analysis, then a significant degree of flexibility exists to ensure that costs will not be prohibitive to the imposition of obligations in overall terms. The lesson from the blocking injunction then is that, where Internet user rights are not necessarily in play or affected to a significant degree, the modality is more likely to be accommodated. If that is the case, then it is user and not OSP commercial fundamental rights that may offer surer footing in resisting the creep of obligations going forward.