Trying to understand Article 13

Kat on a platform

As IPKat readers know, next week the European Parliament will vote on the draft Directive on copyright in the Digital Single Market.

This blog is happy to host the following analysis of one of the most debated provisions in that draft legislative instrument, ie Article 13, as offered by Katfriend Prof Dirk Visser (Leiden University). The text of the full analysis is available here.

Here's what Dirk writes:

"Introduction 

Article 13 is the most important and most controversial provision of the proposed EU Directive on Copyright in the Digital Single Market on which the EU member states and institutions have reached agreement in principle in February 2019. The final vote will be in the European Parliament in late March 2019. 

My analysis - available here - focuses on what the idea of Article 13 is and how it might work out in practice. Which legal issues will arise, which preliminary questions will be put to the CJEU? It is not meant to argue for or against Article 13. The article itself is taken as a given. This paper is merely a first attempt to look ahead and find out what could happen in practice. 

It is obvious that Article 13 is a compromise between many interests and wishes and that is quite contradictory at some points. But that in itself is nothing new. Much of the EU legislation consist of partly contradictory or incomprehensible clauses, because compromise apparently was the only option for reaching consensus. Consequently, it is often up to commentators, practitioners and judges to make the best of it. 

What is the aim of Article 13? 

The aim of Article 13 is twofold, but the target is the same: the large ‘online content sharing service providers’ (‘OCSSP’s’), notably Google (YouTube) and Facebook. The first aim is to make these OCSSP’s pay to rightholders, on the basis of content licences, to the extent rights owners are willing to give them licenses. This would diminish the so called ‘value gap’ between the amount of money the OCSSP’s make out of advertising in and around the content uploaded by their users (a lot of money) and the amount of money rights holders get in return (not enough money). The second aim is to make those OCSSP’s take down, block and filter all content for which the rightholders are not willing to grant a license to such platforms, but which users nevertheless (try to) upload onto these platforms. 

The main concerns are

1. that rightholders will not be willing to license these OCSSP’s, or not on reasonable terms, 
2. that the blocking and filtering will amount to abuse by rightholders and fraudsters, 
3. that it will lead to over blocking of legitimate use, involving all kinds of free speech issues, and 
4. that it will be very costly for the OCSSP’s. For smaller and non-commercial OCSSP’s it might even be prohibitively costly, thus making it impossible for them to stay on the web, or for new OCSSP’s to enter the market. 

The many recitals and subsections of Article 13 try to cater for all these concerns. The paper comments on all these subsections. 

What will happen? 

The Collective Management Organisations (CMOs) and other rightholders that have a mandate to do so, will contact the main OCSSPs, Google and Facebook, in order to negotiate a licensing agreement. And as far as music is concerned, they will probably be able to reach an agreement. For any other kind of works and rights, it is very hard to predict whether licenses will be agreed upon. But CMO’s will try to acquire the relevant rights and try to provide licenses. 

The rightholders that do not want to authorise any use by the OCSSPs, will provide them with ‘the relevant and necessary information’ to enable them to filter and block their works form the platforms. 

The OCSSPs themselves will formally invite everyone and anyone to provide them with a licence. 

The OCSSPs might choose to start filtering everything for which they receive ‘the relevant and necessary information’ and not try to make any effort to decide in advance whether anything is actually protected or not covered by exceptions or limitations to copyright. They could counterbalance this by accepting or reinstating the upload as soon as an uploader states that his upload is not protected or covered by exception or limitation. Or they might allow uploaders to state in advance that their upload is not protected or covered by exception or limitation, and balance this by blocking the material as soon as a rightholder states that that is not true. 

The OCSSP’s might argue that these are the only available and suitable solutions in view of the costs, also ‘in the light of the principle of proportionality’. 

Following that, the OCSSPs must allow either the uploader or the rightholder to complain, and have a human make a decision. That decision can then be appealed through an ‘out-of-court redress mechanism’, to be provided by the Member-States. And that decision can then be contested in a regular court. 

Furthermore, it is imaginable that start-ups might try to remain below the thresholds of an annual turnover of EUR 10 million and the number of monthly unique visitors of 5 million, in order to avoid being considered an OCSSP. 

It is quite certain that OCSSPs, and the rightholders involved, should not wait until the date of entry into force of the Directive before organising ‘stakeholder dialogues to discuss best practices for the cooperation’ between them. 

Kat platforms

What can internet intermediaries reasonably be expected to do? 

In the end the courts will have to decide what the OCSSP’s and other providers can reasonably be expected to do. Unfortunately, Article 13 gives very little guidance for this. It rather describes and creates very elaborate contradicting objectives and obligations. 

The issues addressed in Article 13 also show that the concepts of primary and secondary liability seem to get mixed up. Maybe this is inevitable. It is likely that this approach of liability of OCSSPs for copyright infringement will have its spill over effects on all kinds of other internet intermediaries. On the other hand it also is a continuation of approach visible in the case law of the CJEU. 

Since Promusicae (C-275/06, ECLI:EU:C:2008:54) intermediaries have been obliged to cooperate with rightholders. Too strict a monitoring obligation goes too far (Scarlet Extended/Sabam, C-70/10, ECLI:EU:C:2011:771 & Sabam/Netlog, C-360/10, ECLI:EU:C:2012:85). But intermediaries have to take ‘all reasonable measures, provided that (i) the measures taken do not unnecessarily deprive internet users of the possibility of lawfully accessing the information available and (ii) that those measures have the effect of preventing unauthorised access to the protected subject-matter or, at least, of making it difficult to achieve and of seriously discouraging internet users who are using the services of the addressee of that injunction from accessing the subject-matter that has been made available to them in breach of the intellectual property right’ (UPC Telekabel Wien, C-314/12, ECLI:EU:C:2014:192). If intermediaries become too much involved, they become primary infringers with full liability (Stichting Brein/Ziggo, C-610/15, ECLI:EU:C:2017:456). 

There will be much interesting caselaw of the CJEU to come, firstly in the two cases from Germany, YouTube (C-682/18) and Elsevier (C-683/18) . 

There will be a continuing discussion all over the world about what which kind of ‘reasonable’ internet intermediary can ‘reasonably’ be expected to do: pay (share advertising income or other income) and/or take preventive or repressive action regarding copyright infringement, while balancing fundamental rights, especially free speech. The same, probably even more important, discussion will continue regarding hate speech, fake news and any other kind of ‘undesirable’ information."