Document Actions

No section

Proposals from Berlin and Paris – Intermediary Liability in European Copyright Law

  1. Jonathan Griffiths


Two very different proposals on copyright policy – one a privately drafted document, the other a governmental report – are published in this edition of JIPITEC. There is an interesting point of intersection between them because they both consider the difficult question of the liability of online intermediaries for users’ infringements. The first document is “The Berlin Gedankenexperiment on the Restructuring of Copyright Law and Authors Rights”. This is a wide-ranging proposal for a complete recasting of the legal system that promotes the production of, and controls the use of, creative goods. The second policy document has a more limited focus. The French High Council for Literary and Artistic Property (“CSPLA”)’s Mission to Link Directives 2000/31 and 2001/29 – Report and Proposals (“Mission Report”) aims to provide a persuasive intervention in current policy discussions at European Union level concerning the liability or, more appropriately, the non-liability, of online intermediaries for copyright infringement. In this brief introduction, I outline the scope of both proposals and reflect briefly on their recommendations.


1. The Berlin Gedankenexperiment on the Restructuring of Copyright Law and Authors Rights*

The Berlin Gedankenexperiment is the product of a project undertaken by a panel of German experts, predominantly from academic and “new media” backgrounds.  [1] It develops an earlier set of guidelines on copyright policy issued by the “Internet & Gesellschaft Collaboratory”,  [2] which is supported by, amongst others, Creative Commons, Google and the Wikipedia Foundation. The project takes a “blank page approach”, allowing the Gedankenexperiment to escape prevailing copyright norms where appropriate. However, it is clearly shaped by the view that the current system of copyright and author’s rights is problematic in a number of respects.

The dominant problem at which the proposal takes aim is the transferability of legal entitlements in creations from authors to other categories of actor (generally to “exploiters”, under the terminology employed). On this point, it is argued that, where one legal actor steps into the shoes of another in this way, there is a risk that the fundamental purpose of a legal regime designed to foster creativity will be subverted:

“Confusing and mixing authors’ and exploiters’ interests opens space for manipulative arguments, which may foster undesirable developments. These lead to a conflict of values, which may ultimately undermine the copyright law system as a whole.”  [3]

The Gedankenexperiment seeks to avoid this problem through the establishment of strict distinctions between the interests of different legal actors involved in the production and use of creative work. Under the proposed system, different categories of legal actors in the creative process (“authors”, “exploiters”, “non-commercial users” and “intermediaries”) are each accorded their own entirely independent rights and duties. The legal position of each is balanced against that of others, without attribution of “structural superiority” to any amongst them. The underlying idea is that the separate contribution of each to the generation of creative products should be separately recognised and protected.

The authors of the Gedankenexperiment suggest that a whole-hearted commitment to this core idea would produce a legal structure differing from that which is currently applicable. For example, under the system proposed, a creator (such as a novelist) would be able to grant an “exploiter” (such as a publisher) (contractual) permission to exploit a protected work. However, as a matter of law, such permission could only be granted for a limited period of time.  [4] At the same time, the publisher would itself acquire its own separate legal right in its published edition (recognising its own distinctive contribution to the dissemination of creativity). At the end of the limited period of permission, the publisher could continue to market its own published editions. However, from that point forward, it would potentially be subject to competition from other published editions permitted by the author.

The proposal is not intended to establish a fully codified body of rules. It is, after all, a Gedankenexperiment and it therefore raises, but does not come to a concluded view on, a number of features of the proposed system - including the precise duration of the various forms of protected interest and the specific treatment of complex works such as films and of works created by employees. In tracing the outline of a legal structure in this way, the project insulates itself from detailed critical analysis. Nevertheless, even against this avowedly sketchy background, it invites questions from the concretely-minded. For example, one of the categories of legal actor to which the proposal attributes rights and duties is described as “non-commercial users”. Such users have a right to carry out acts which either (i) fall within a specified catalogue of use rights or (ii) are covered by an open fair use-type norm. However, the document makes no mention of commercial users in this context. Such users presumably fall within the category of “exploiters”, who have their own designated duties and rights. However, while the proposal traces the entitlement of exploiters who have been granted contractual permission to use a copyright work, it does not appear to deal explicitly with the situations in which a commercial actor is typically entitled to use a copyright work without permission under current law (for example, for the purpose of quotation, news reporting or parody). This must surely simply be an omission. If the project team had intended to restrict the circumstances under which commercial users are entitled to commit otherwise infringing acts, it would surely have done so explicitly.  [5]

The Gedankenexperiment’s ‘blank page” approach undoubtedly brings a breath of fresh air to the sometimes poisonous debate on copyright reform. However, while it might appear to be based on a radical premise, its recommendations are relatively incremental in some respects. Many features of existing copyright and authors’ rights systems – including creators’ moral rights, copyright contract regulation and the special regimes applicable to film productions and to creations by employees – are retained. Indeed, even those elements of the proposal involving significant change to the existing legal order (with the possible exception of the elaborated “balance of independent rights” system outlined above) echo suggestions for reform made elsewhere in the recent past. Thus, for example, under the Gedankenexperiment, the terms of protection for authors’ and exploiters’ rights would be significantly shorter than those currently prevailing in European and international law. Many such calls to reduce the term of copyright so that it more closely reflects its underlying rationales have been made. Similarly, the proposal’s suggestions that authors’ promises of exclusivity should be noted in a public registry, that the continuation of the exclusive protection for those that invest in the dissemination of works should be conditional on the payment of progressively increasing fees and that exceptions and limitations should be replaced with a set of “user’s rights” for non-commercial users and a fair use-type clause all are not without precedent. Indeed, in the last case, recent judgments of the Court of Justice may already have delivered such an outcome in the European Union.  [6]

Even the Gedankenexperiment’s most distinctive proposal, the “balance of independent rights” system may not have been breathed into life ex nihilo. To this common lawyer’s untrained eye, the project’s emphasis on the non-transferability of author’s entitlements looks rather like a super-charged extension of the current German system for the protection of author’s rights. It would appear that the page upon which this stimulating proposal has been drawn up may not have been entirely blank after all.

2. Intermediaries

The relatively reasonable, modest characteristics of the Gedankenexperiment are also apparent in its proposals concerning the potential liability of online intermediaries for copyright infringement. This is one of the most contested questions in current debates on copyright policy. Legal systems have struggled to develop appropriate theories to impose responsibilities on intermediaries without over-burdening them in a manner that would unreasonably hamper the functioning of new forms of communication technology. Considerable uncertainty on this question persists in many jurisdictions.  [7] Within the European Union’s legal order, online intermediaries benefit from the E-Commerce Directive’s “safe harbours” for information service providers.  [8] Under these general provisions, when information service providers function as “mere conduits”  [9] , “caches”  [10] or as “hosts”  [11] for information originating from others, they enjoy immunity from liability for damages for, inter alia, copyright infringement as long as certain conditions are satisfied.  [12]

Thus, for example, under Art 14 of the E-Commerce Directive, where an information society service provider stores information provided by a recipient of its service (i.e. it functions as a “host”), it will not be liable for damages if it (i) has no actual knowledge of illegal activity and is not aware of facts or circumstances from which illegal activity is apparent and (ii) acts expeditiously to bring any illegal activity to an end on receiving such notice.  [13] The scope of this provision is contested and some have argued that, while it might have been appropriate to grant such an immunity in the early years of development of networked electronic communications, online platforms, such as YouTube, now make vast profits through the hosting of unlicensed copyright materials posted by users and have no need for such shelters from liability. Critical concerns have been exacerbated by the Court of Justice’s broad interpretation of Art 14 in cases such as Google France, in which the Court interpreted the hosting immunity as applying in circumstances in which a service provider lacks specific knowledge or control of stored data and fulfils a “merely technical, automatic and passive” role.  [14] The European Commission is currently considering this issue within its review of the Union’s copyright rules.  [15]

The Gedankenexperiment advocates a nuanced approach to the legal responsibility of intermediaries:

“For an appropriate balance of interest, it seems necessary to differentiate between intermediaries whose offers tend to compete with goods and services from exploiters/creators or may even substitute them (“competing intermediary services”), and those whose offers complement the goods and services of rights owners or even make them possible in the first place (“complementary intermediary services”).’  [16]

Intermediaries that are “close to the content” (such as video and image hosting platforms) are considered to be more likely to compete with the offering of a creator/exploiter than intermediaries that are “far from the content” (such as, for example, technical internet access providers). Within the category of “competing” intermediaries, the Gedankenexperiment makes further distinctions. It recognises, for example, that the full range of legal remedies ought to be available against an intermediary that is concerned purely to freeride on the creative contribution of authors and exploiters. However, it is acknowledged that the situation of other online platforms is more ambiguous because, while they may cause prejudice to rightholders’ distribution channels, they also promote public welfare in certain important respects.

In keeping with the approach that it adopts throughout, the Gedankenexperiment suggests that, in such circumstances, the interests of the various affected categories of actor must be balanced and that:

“A possible result of such a balancing act could be that providers of legitimate (potentially) competing offers…would be given an obligation to pay monetary compensation in lieu of their users, e.g. in the form of shares in revenue or an adequate compensation.”  [17]

Under such a system, which is acknowledged to bear similarities to some currently applicable mechanisms, an intermediary would have to decide itself whether to pass on the costs of such compensation to users or to finance the payment in other ways (presumably, for example, by advertising). In return for the assumption of an obligation to pay compensation, intermediaries would be completely relieved of monitoring obligations  [18] and would be provided with immunity from liability for its users’ infringements.

In truth, this analysis of the problem does not get us particularly far. On the face of it, the “close to the content”/”far from the content” distinction is more descriptive than analytical and the suggestion that the potential loss of revenue to rightholders could be made up through a balanced compensation system is not revolutionary. Nevertheless, the Gedankenexperiment’s strong commitment to the recognition and reconciliation of competing interests takes the notion of “balance” beyond rhetoric and, at least, establishes a foundation for the exploration of the problem of intermediary liability. By contrast, the second proposal published in this edition of JIPITEC addresses the same question but takes a very much less tentative and reflective position.

3. The Mission to Link Directives 2000/31 and 2001/29

The High Council for Literary and Artistic Property (Conseil supérieur de la propriété littéraire et artistique, CSPLA) is responsible for advising the Minister of Culture and Communications of the French Republic on matters relating to literary and artistic property. Created under legislative order, it has produced a number of reports on questions relating to authors’ rights.  [19] Its “Mission to Link Directives 2000/31 and 2001/29”  [20] was presided over by Professor Pierre Sirinelli  [21] and reported at the end of 2015.  [22] The Mission, which consulted a number of stakeholders,  [23] focused on two questions. First, it considered whether: “…[T]he regimes implemented by Articles 12 to15 of the E-Commerce Directive of 8 June 2000 (Directive 2000/31/EC) truly provide a full understanding of the activities of certain service providers (Web 2.0 in particular) who were barely in existence when this legislation was adopted?”

Secondly, should the answer to the first question prove to be negative, the Mission’s role was to investigate potential solutions to the problem presented by the inappropriate application of the E-Commerce Directive’s safe harbors in the current technological context. The Mission Report builds on an earlier CSPLA report, likewise led by Professor Sirinelli, on proposals to revise the Information Society Directive.  [24] That earlier report recommended that the E-Commerce Directive’s immunities should be re-examined because of their negative effect on the holders of rights in literary and artistic property.

In these circumstances, it is perhaps not surprising that the first question (“Does something need to be done?”) does not detain the authors of the Mission Report for very long. They note a near unanimous view among stakeholders that platforms’ claims to immunity are problematic.  [25] The Mission then goes on to consider the cause of, and potential solutions to, the problem that it has identified. It is highly critical of the Court of Justice interpretation of the scope of the Art 14 immunity as covering the activities of highly profitable platforms (described as “false hosting providers” in the report). The finding, in Google France, that information service providers fall within the safe harbour where they do not play an active role, so as to give them knowledge of, or control over, stored data, is characterized as an error of interpretation that should be remedied through European legislation. The Mission’s preference in this regard is for the implementation of a copyright-specific solution rather than for a general revision of the E-Commerce Directive.

More specifically, it recommends the introduction of a new Article in the Information Society Directive (Art 9a):

“Without prejudice to Articles 12 and 13 of the Directive on electronic commerce, information society service providers that give access to the public to copyright works and/or subject-matter, including through the use of automated tools, do not benefit from the limitation of liability set out by Article 14 of said Directive.

These service providers must obtain permission from the relevant rightholders as they, either alone or with the participation of users of their services, are implementing the rights set out by Articles 2 and 3.

Such permission covers acts performed by users of their services when they send the copyright works and/or subject-matter to the aforementioned service providers in order to allow the access set out by sub-paragraph one, as long as these users are not acting in a professional capacity.”  [26]

According to the Mission Report, this provision would restore a “better sharing of value” by distinguishing between service providers which purely “store” information for third parties (and would presumably still be covered by Art 14) and service providers which “give access to the public” to copyright works, and other protected material, and which would therefore be liable for infringement (with their infringing users).

Such a change would diminish the scope of the Art 14 immunity as it is understood today. Given that the Report’s aim is to amend the Information Society Directive, it also seems possible that the introduction of the new Art 9a might risk expansion of the scope of liability well beyond the “YouTube”-type situation at which the Mission is ostensibly aimed. If the draft provision were introduced, the Court of Justice would have to wrestle with the relationship between the concepts of “giving access to the public”, “making available to the public” and “communication to the public”. As a result, it seems likely that the introduction of the proposed Art 9a would add complexity to an already confused area of jurisprudence.  [27] As any shifts in the current situation would be likely to favour right-holders, this consequence might not be entirely unwelcome to the authors of the Mission Report. Nevertheless, they acknowledge that a diminution in the scope of Art 14 might cause difficulties for online intermediaries. Their proposed solution is, first, the introduction of transitional protection for intermediary business models developed on an expectation of immunity and, secondly, the implementation of a “duty of collaboration” between rightholders and service providers. It is perhaps rather ironic that, while the Mission Report is based on the assumption that the initially intended reach of the E-Commerce Directive’s immunities is no longer appropriate in current technological conditions, the solution that it identifies is a reversion to legal orthodoxy, anchored by the authority of the Berne Convention,  [28] a Treaty first agreed in 1886 and last revised in the 1970s.

By contrast with the Gedankenexperiment’s somewhat incomplete and speculative tracing of principle, the Mission Report is pragmatic and detailed. It is therefore not surprising that its faults are very different from those of the open-minded Gedankenexperiment. The disdainful rhetoric of the report leaves a reader with the strong impression that it might have been possible to predict the broad thrust of its conservative recommendations in advance of its consultation with stakeholders. It takes no account whatsoever of arguments that could be advanced in favour of a less conservative solution to the “value sharing” issue. Presumably, one of the underlying reasons for the Court of Justice’s broad interpretation of Art 14 in Google France was its sense that some of the public benefits of technological advance might be lost if right-holders were granted unmitigated dominion over the activities of online platforms. The Mission Report does not engage with such concerns. Similarly, there is no mention of the fundamental rights framework upon which the Court of Justice has structured all its recent responses to intermediary liability. Through the application of the Charter, the Court has acknowledged the need to balance the right of property of copyright owners with the right to conduct a business of service providers and the right of freedom of expression, and access to information, of users.  [29] In the Mission Report, no time is wasted on a discussion of this framework of competing rights or, indeed, on the due process rights of users under the “notice and take down” process facilitated by the E-Commerce Directive’s safe harbours.

It is impossible to escape the conclusion that the Mission’s predominant intention was to put down a marker for current discussions on copyright reform at European level. However, it seems unlikely that the European legislator will be entirely persuaded by its call to apply a right-maximalist form of regulation in the online environment. The “duty to collaborate” that the Mission envisages is surely too weak to offer adequate protection to online intermediaries (and, therefore, to the public interests that their activities support).  [30] The strengthening of such a system based upon forced negotiations might, however, bring the proposal closer to the zone occupied by the Gedankenexperiment’s suggestion of an obligation to pay compensation or, indeed, to the current legal situation in which platforms pay a proportion of advertising revenues to creators and have negotiated licence agreements with bodies representing right-holders. I would argue that a half-appropriate solution to the problems presented by platforms’ hosting activities is much more likely to be found through a difficult exploration of this contested zone than through any reassertion of doctrinaire copyright orthodoxy.

* Jonathan Griffiths, BA (Oxon) MA is a Reader in Intellectual Property Law at Queen Mary University of London.

[1] The German original is available at - The English translation is reprinted on p. 76 of this volume.

[3] Berlin Gedankenexperiment, 3.

[4] Ibid, 4. The document does not provide a final recommendation of an appropriate period for which permission may be given. – It should be noted that in its recent proposal for a law improving the claim of authors and performers to adequate remuneration, the German Federal Government proposes, albeit on the basis of a different legal construction, a somewhat similar result in providing that an author, who has granted an exploiter an exclusive exploitation right against payment of a lump sum fee, shall be free after ten years to exploit his work otherwise, with the initial exploiter retaining a non-exclusive right to continue his own exploitation; see § 40a (1), BT-Drucks. 18/8625.

[5] Editor's note: The drafters of the Gedankenexperiment have explained that the term „non-commercial user“ was only chosen to set a clear terminological distinction between users who are entitled to use protected material by statute and those who need a license for their uses (the latter are referred to as „exploiters“). This does not necessarily mean that there uses for a commercial purpose cannot fall under the user’s statutory rights. Whether this is the case or not, depends on the particular balance of interests.

[6] Through its reliance on the Charter of Fundamental Rights in interpreting the copyright acquis. See, for example, (C-201/13) Deckmyn v Vandersteen, 3rd September 2014.

[7] For discussion, see, for example, J Wang, “Not all ISP Conduct is Equally Active or Passive in Differing Jurisdictions: Content Liability and Safe Harbor Immunity for Hosting ISPs in Chinese, EU and US Case Law” [2015] EIPR 732; Federal Supreme Court (Bundesgerichtshof); 26 November 2015 – Case No. I ZR 174/14, “Germany: Disturber Liability of an Access Provider” [2016] IIC 481; A Gärtner & A Jauch, “Gema v RapidShare: German Federal Supreme Court Extends Monitoring Obligations for Online File Hosting Providers” [2014] EIPR 197; C Angelopoulos, “Beyond the Safe Harbours: Harmonising Substantive Intermediary Liability for Copyright Infringement in Europe” [2013] IPQ 253; M Leistner, “Structural Aspects of Secondary (Provider) Liability in Europe” [2014] Journal of Intellectual Property Law & Practice 75.

[8] 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.

[9] Art 12.

[10] Art 13.

[11] Art 14.

[12] The provisions do not preclude the grant of injunctions against service providers conducting the specified activities (Arts 12(3), 13(2), 14(3). Information service providers also benefit from a prohibition on the imposition of a general obligation to monitor for infringement and/or a general obligation actively to seek facts or circumstances indicating illegal activity (Art 15).

[13] Art 14(1)(a), (b).

[14] (C-236/08 – 238/08) Google France v Louis Vuitton [2010] ECR I-2417. See also (C-324/09) L’Oréal SA v eBay [2011] ECR I-6011.

[15] See European Commission, Towards a Modern, more European Copyright Framework, 9th December 2015, COM (2015) 626 final.

[16] Berlin Gedankenexperiment, 13 (footnote omitted).

[17] Ibid, 14.

[18] While Art 15 of the E-Commerce Directive precludes general monitoring obligations, some Member States have sometimes imposed more specific monitoring obligations on information service providers. See, for example, A Gärtner & A Jauch, “Gema v RapidShare: German Federal Supreme Court Extends Monitoring Obligations for Online File Hosting Providers” [2014] EIPR 197.

[21] Université Paris-I (Panthéon-Sorbonne).

[22] 3 November 2015. Vice-Presidents of the Mission were Josée-Anne Benazeraf (lawyer at the Paris Bar) and Alexandra Bensamoun (Senior Lecturer, Université Paris-Sud).

[23] Although “some technical service provider representatives opted not to respond to the mission's invitation” (Mission Report, 2).

[24] Conseil supérieur de la propriété littéraire et artistique, Report of the Mission on the Revision of Directive 2001/29/EC on the Harmonisation of Certain Aspects of Copyright and Related Rights in the Information Society, December 2014.

[25] “This affirmative response would have been unanimous but for the caution of certain technical service providers” (Mission Report, 3).

[26] The proposed new Article is accompanied by two proposed new recitals (16a and 24a).

[27] For recent interventions on this issue, see (C-160/15) GS Media (Opinion of AG Wathelet, 7th April 2016); (C-117/15) Reha Training (Court of Justice, Grand Chamber, 31st May 2016).

[28] See Mission Report, 11, 12.

[29] See, for example, (C-275/06) Promusicae v Telefonica de Espana [2008] ECR I-271; (C-70/10) SABAM v Scarlet Extended [2011] ECR I-11959; (C-314/12) UPC Telekabel Wien GmbH v Constantin Film Verleih GmbH, 27th March 2014. For discussion, see C Angelopoulos, “Tracing the Outline of a Ghost: the Fair Balance between Copyright and Fundamental Rights in Intermediary Liability” (2015-16) Info – the Journal of Policy, Regulation and Strategy for Telecommunications, Information & Media” 72.

[30] The Mission Report has not been particularly well received by online intermediaries. See CSPLA, Commentaires des organismes professionnels membres du CSPLA sur le rapport relatif à l'articulation des directives 2000/31 et 2001/29, 2-5 (Response of ASIC, l’Association des Services Internet Communautaires). For an example of an interpretation of copyright rules in a manner that recognises to develop legal principles in the face of technological change, see the recent Opinion of AG Spuznar in (C-174/15) Vereniging Openbare Bibliotheken, 16th June 2016.



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

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