radiobruxelleslibera

[continuing from the previous post about the link-tax]

Art. 17 (former 13) of the Copyright Directive, concerning the liability of online content-sharing platform and the upload filters, is the most “systemic” part of the European copyright reform.

The primary liability of content-sharing platforms 

Article 17 creates a primary liability for online content-sharing service providers giving public access to copyright-protected works uploaded by their users. This is about Youtube or Youtube-like companies. Primary liability means that the platform is liable ipso facto when the illicit content is uploaded, because of the simple act by a third person (the platform user and uploader). It is a very burdensome liability operating ex ante, similar to the ones the civil law provides for very dangerous activities. To better explains this legislative innovation, please note that the current liability regime (in force since 2001 with Directive 2000/31) operates ex post, i.e. via the removal of the illicit content after the uploading. There is a huge difference between a) imposing a system preventing the upload of content ex ante, on one side; and b) providing an ex post removal regime, on the other; and legal implications are very different. If you catch this difference, you are looking at the core of the Copyright debate. All the rest is fresh water 🙂

As a consequence of the above “ipso facto” primary liability, content-sharing platforms must be authorized to have their content uploaded and made available on their website. If not, they are liable and should pay damages. But how can you be authorized even before users decide to upload a given video or music? You should foresee the future. This mechanism seems contrary to common sense and in fact it is.

How could content sharing platforms avoid primary liability?

Still, art. 17 provides that Youtube & CO. can avoid liability thanks to the so-called “mitigation measures”, consisting in demonstrating that they;

(a) made best efforts to obtain an authorisation, and

(b) made, in accordance with high industry standards of professional diligence, best efforts to ensure the unavailability of specific works and other subject matter for which the rightholders have provided the service providers with the relevant and necessary information; and in any event

(c) acted expeditiously, upon receiving a sufficiently substantiated notice from the rightholders, to disable access to, or to remove from, their websites the notified works or other subject matter, and made best efforts to prevent theirfuture uploads in accordance with point (b).

Filter or not filter?

Point b) is obscure. What does it mean “best efforts to ensure the unavailability of content on the platforms”?  Does this provision imply mandatory upload filters or not? This is a tricky. The new Copyright Directive never explicitly mentions filters and its supporters, both stakeholders and persons of the European institutions, have been vocal in declaring  that upload-filters are not foreseen by the provision. See for instance rapporteur Axel Voss:

Sometimes

However, the sole available technological instrument that a platform can use to effectively avoid primary liability, that is to say to anticipate and prevent a user from uploading and making public illicit content, is a un automated upload-filter. Only filters can enable platforms to identify and block posting and -reposting of illicit content, before they become public.  No other technological instruments are available, since non-automated instruments, such as for instance a human check, would be unfeasible. Therefore, only two options are possible:

a) either we accept, even obtorto collo,  that only (automated) upload filters may preventively and effectively block illicit content, so that to give a sense, although controversial, to this Copyright Directive; or

b) we could be satisfied with any solutions other than automated upload filters. But this means that such solution could operate only ex post, after illicit content is uploaded and made public (similarly to the current liability regime). Tertium non datur.

Why this Directive pushes towards filtering, but with vague and imprecise terms?

Omitting the term “uploadfilter” in the Directive was a clear and intentional political decision, necessary  to facilitate the approval, this is why the supporters of the Directive have been denying that the same is mandating or implying upload filters. At the same time, the Directive itself requires platforms to make make “best efforts” to block illicit content, an objective which could be realistically achieved only by applying upload filters. In practice, the Directive aims at something specific which, however, cannot be said and written down. In the best scenario, it looks like the directive was written down by rabbits, not persons. In the worst scenario, it looks like an imbroglio.

National authorities to decide about filtering?

Since art. 17 is so vague, it will be a matter of national implementation and interpretation by national authorities. When looking at practical cases, national authorities will be able to take various directions but, at the end, only the same 2 options are available: either mandating upload filters, as the only effective instruments to prevent ex ante illicit content to be uploaded on platforms; or any other solutions which, however, will basically work on a ex-post control basis.

What about voluntarily filtering? This could be a possible solution but, as far as filters are not mandatory, they could be replaced by “best efforts” ex-post mechanisms. And, as we know, filters are a big investments that only big companies like BigG can afford. “Simple” filters could be found in the market but would be ineffective and overblocking, with the risk to kill the business.

What are national authorities expected to do in practice? If they look at the text of article 17, imposing an express upload-filters provision will not be obvious, also considering that the Directive contains another contradicting obligation (art. 17(8)): “The application of this Article shall not lead to any general monitoring obligation“.

Rights-holders will probably complain because they pretend unauthorised content to be effectively blocked ex ante, but then they also should explain why upload filters were not required before voting of the directive, while after the approval they are.

Anyway, there might be national governments which already have clear ideas. The French government, for example, just the day after the approval of the Directive officially stated that they are going to adopt upload filers because they are the only way to make art. 17 working:

Will the Commission recommend filtering?

Unless you are French, it will be difficult for national authorities to take a straight decision on the topic. It is more likely that vagueness and contradictions of the Directive will be transposed into national law, without solving the problem which, therefore, will have to be dealt on the ground: at the very hand, a national judge will have to take the decision on the basis of vague provisions and therefore jurisprudence could strongly differ from country to country. A preliminary ruling to the European court of justice could also be possible.

In the meanwhile, the European Commission will be required to provide guidance on the matter after having heard the views of all concerned stakeholders (platforms, rights holders, users, ecc.) gathering together (art. 17(10)). It is unlikely that such stakeholders will magically find a an agreement on the filters subject, therefore at the very end the Commission will remain holding the bag and will have to decide what to do, what to recommend. Remarkably, the entire system seems designed to postpone, as longest as possible, any decision.

Worth-noting, next European Parliament and Commissioners may be a bit colder vis-à-vis copyright supporters. Therefore, will next Commission recommend national authorities to mandate upload filters or could, instead, recommend a more soft interpretation? This could probably become a question that the new European Parliament will ask to the next candidate commissioner for the Digital Agenda, in July 2019.

The new Copyright Directive have been approved by the European Assembly with an important but not large majority: 348 in favor, 274 against, 36 abstention. Quite an important number of MEP avoided the voting, probably because they fear the revenge by citizens at next European elections in May. While the PPE (christian democrats) voted massively in favor of the Directive, other political parties , in particular S&D (socialists) and ALDE (liberals) were basically split.

The completion of the legislative process

The path for the approval and effective application of the directive will still be long, although the concerned provision cannot change anymore. Council is expected to endorse very soon the same text approved by the Parliament (likely in April 2019) which then will be published in the Official Journal. At the point the legislative process will be duly completed. However, being a directive, the text still will need to be implemented into national law, and Member States will have 24 months time. Therefore, the concerned provisions are expected come into force in mid-2021.  

Implementation will not be an easy matter: especially with regard to the most controversial provisions, such as article 11 (now 15 with the new numbering) about the so-called linktax, and article 13 about liability of videosharing platforms and use of uploadfilters (now art. 17 with the new numbering) are written sometimes in a very vague way. The vagueness is not result of technical incapacity, rather it is the consequence of the difficulties in finding an acceptable compromise for everyone. Bismark’s saying whereby “Laws are like sausages. It’s better not to see them being made” is still valid more than ever.

As a final result, the new Copyright Directive risks to be implemented, interpreted and applied in many way from country to country. European Guidelines will probably facilitate more harmonization, however they will not be binding for national authorities.

The ancillary copyright 

According to Art. 15 of the Directive (former art. 11) publishers  are granted a new ancillary copyright (pursuant to the Copyright Directive 2001/29/EC) for the online use of their press publications by “information society service providers”. The ancillary right will last two years after the press publication is published (art. 15(4)).

Remarkably, art. 15 will apply to any Internet service provider, with no exceptions about turnover, size and scope of the activity ecc. Being Google, a start-up, a charity or or an educational entity will be the same: there is no exemption (there might be an exemption just for persons, see below). This is the reason why Wikipedia has been so vocal against the reform. Someone believed that Wikipedia was exempted, but that exemption concerns just videosharing, not publishing rights (see art. 2(6)). Even a porte-parole of the European Parliament got confused on this matter:

Surprise: the ancillary right is still a link-tax

According to art. 15(1)3, the ancillary right should nor apply to acts of hyperlinking, neither upon individual words or very short extracts of press publications (art. 15(1)4). What does this mean? The ambiguity of the provision is the reason why the name “link-tax” remains valid despite the attempts to remove it. The clarification that hyperlinks are not taxed per se is good but insufficient, since the social and economic value of hyperlinks derive from combination with words and images referring to the linked content. Should such words be more than one (at least 2) or form an extract which is not considered “short”, then the link-tax (or whatever is the name) will apply. In particular, it will not be possible to express with the hyperlink the title of the linked press article.

The only way to be safe would be to use hyperlinks combined with words disconnected or irrelevant from the content they refer to. However, this is not the way hyperlinks and the Internet work.

The way this provision will be applied from country to country may strongly vary. This the reason why Wikimedia Italia is so worried. In an exchange of tweets with a journalist of La Stampa, their worries are pretty clear:

Will normal people be exempted? 

According to art. 15(1)2, the ancillary right should not apply to private and non-commercial uses by individual users. This is commendable, however the factual application may be very limited, since pure “Private or non commercial use” in the Internet is just theoretical. Internet services (blog, social platforms) normally require users to accept a set of contractual conditions which include eventual economic exploitation of content (mainly via revenue sharing of online advertising). Therefore, in order to make the exception valid, all Internet services should re-write their Terms & Conditions in order to clearly separate “private/non commercial” from “commercial” usage (even when commercial is just potential). This separation, however, may be burdensome and incompatible with current business models and, so far, no-one made an assessment to verify whether it can be made without affecting normal usage.

In any case, Internet providers will not be obliged to create such separation amongst their subscribers, with the consequence that the latter will remain in a limbo: private or commercial use? Who knows.

Will Google finally pay?

A part from the above, the greatest paradox is that the real objective of this fight, namely Google, may easily avoid the payment. The level of the fee payable with link-tax can be negotiated or even waived, therefore Google, thanks to its economic strength in the online world, could easily get the better conditions, or even pay ZERO as it already happened in Germany. Smaller Internet provider will not have the same privilege, though. Of course, one could start an antitrust action against Google for abuse of dominant position (in Germany someone did it), however Google is not dominant in the news aggregation market, therefore the action will likely fail.

At least, journalists will be paid more?

According to art. 15(5), journalists should receive an “appropriate” share of the revenues created by the link-tax. The provision does not provide further details and therefore it is expected to be applied pursuant to the willingness of publishers. This is a bad news for journalists who maybe had dreamed to be better paid.

[Another article on the upload filters to follow] 

Attenzione! …. for those who perform acts of piracy from the Internet connection at home. The fact of residing in a home with other roommates, all potentially capable of using the Internet, will not be enough to exempt themselves from liability by claiming, as a defense, that offender could be someone else. The Court of Justice of the European Union has ruled that there must be a way to allow the a copyright holder to defend his interests in case of violation perpetrated through shared Internet connections.

It may seem like a common sense solution but, in the case originated from Bavaria, the German rules on the protection of family life did not allow further investigations within the family unit that used the offending Internet connection. In fact, a German publisher, Bastei Lübbe AG, had sued a Bavarian citizen, Mr. Strotzer, because through the latter’s Internet connection had been downloaded, and subsequently shared on a peer-to-peer platform, a file containing an audiobook of the publisher. Mr. Strotzer defended himself by denying having infringed the copyright of the publisher and stating, moreover, that his parents, with him cohabiting, had equal access to the connection, without however providing further clarification on the possible use that the parents themselves would have made of the Internet connection. The Bavarian court of first instance rejected the application of Bastei Lübbe, considering that the fundamental right for the protection of family life prevailed in this case. However, the appeal judge felt differently and asked the European court whether such a defense may be sufficient to exclude the responsibility of the holder of the Internet connection.

The European court simply stated that right holders must have an effective form of redress or tools to enable the competent courts to order the disclosure of the necessary information. It is therefore not a question of weakening the fundamental right to privacy, but rather a signal sent to the German authorities to provide all the necessary instruments to ensure that there is a balance between the various interests at stake, including protection of intellectual freedom. In the present case, Mr Strotzer will therefore have to argue better about the use of his Internet connection by third parties, including mother and father.

The case of shared Internet connections can go beyond home and involve broader situations, such as the communities of students, workers, friends, as well as public WiFi connections. The European ruling does not oblige the holder of the Internet connection to ensure the identification of each user, rather to make themselves more cooperative with the courts in the search for offenders. The European court evokes a possible remedy, namely an objective liability of the Internet subscriber (as it happens with cars), but this seems to be an extreme solution, to be used only when the national legislator does not allow operations to identify the offenders among those who have access to a shared Internet connection.

The present case may seem a little excessive given the family context from which it originates, but it must be borne in mind that the objective of the judicial action was not the illicit use of a protected content (a condemnable action but with a modest impact on the publisher), rather the uploading of a file on a peer-to-peer platform accessible to anyone.