The EU Copyright Directive: the three most controversial provisions

In April 2019, after years of debate and a rather controversial EU legislative process, the Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market (hereafter referred to as the Copyright Directive) was finally adopted. The purpose of this directive was to adapt copyright protection to the digital age in order to create an internal market for digital content and services. Three specific targets were set: (1) to adapt exceptions to reflect the digital and cross-border environment (2) to ensure wider access to content, and (3) to achieve a well-functioning marketplace for copyright.

In April 2019, after years of debate and a rather controversial EU legislative process, the Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market (hereafter referred to as the Copyright Directive) was finally adopted. The purpose of this directive was to adapt copyright protection to the digital age in order to create an internal market for digital content and services. Three specific targets were set: (1) to adapt exceptions to reflect the digital and cross-border environment (2) to ensure wider access to content, and (3) to achieve a well-functioning marketplace for copyright.

With less than one year left for member states to implement the Copyright Directive into national law, controversy and uncertainty surrounding some of its provisions remain. Three important sets of rules are particularly hotly contested: the text and data mining exceptions (Articles 3 and 4), the press publishers’ right (Article 15), and the liability of user-upload platforms (Article 17).

1) The text and data mining exceptions

Articles 3 and 4 of the Copyright Directive provide two mandatory exceptions to copyright infringement for text and data mining (TDM).

TDM is defined as any automated, electronic analysis of large amounts of data in order to extract information and patterns that cannot be processed or detected by human reading. These types of data processing are increasingly important in various fields, including not only scientific research and journalism but also commercial uses in a consumer context. In addition, artificial intelligence often relies on some sort of data mining and analysis.

From a copyright perspective, TDM reproduces data and the content embodying them in order to carry out its analysis. Such copying, irrespective of its purpose, permanence or transient character, could be considered as an act of exploitation and thus a copyright infringement.

The new Copyright Directive provides two specific exceptions, one allowing TDM for scientific research, the other allowing it for any other purpose provided more stringent conditions are met.

The first exception (in Article 3) is for acts of TDM for the purpose of scientific research carried out by research organizations and cultural heritage institutions. A research organization is defined as either a not-for-profit entity or an entity tasked by a member state with a public interest mission. A cultural heritage institution is defined as a publicly accessible library or museum, an archive or a film or audio heritage institution. These definitions therefore exclude commercial research institutes and public broadcasting organizations, which are not covered by Article 3.

Research or cultural heritage institutions should obtain lawful access to the materials used for TDM and they should store copies of such works with an appropriate level of security.

The second exception (in Article 4) permits TDM for any other undefined purpose, and this therefore covers TDM carried out for commercial or for-profit reasons. It allows everyone with lawful access to the works in question to conduct TDM, but only as long as the right to carry out TDM has not been expressly reserved by the right holders. Right holders are therefore entitled to reserve the use of their works either by legal means (such as contracts, general terms and conditions or a unilateral declaration) or by technical means (like the prevention of indexation by search engines).

This opt-out means that content owners, such as, for example, scientific publishers, could monetize (by granting licenses) or even prohibit for-profit TDM.

The exception thus relies on the “goodwill” of the right holders. If technical measures are put in place to prevent commercial data mining, this could result in researchers being increasingly deterred from performing TDM, and this could in turn have a negative effect on research and innovation in the EU. Furthermore, the requirement to lawfully obtain access to the documents and data could become an obstacle for anyone wishing to benefit from this exception. For example, a journalist who obtains documents from a whistleblower is unlikely to be able to rely on the exception.

2) The press publishers’ right  

Following the fierce (mainly national) debates between Google and press publishers, and induced by the significant power imbalance between both players, Article 15 grants press publishers based in the EU a new (ancillary) online IP right. Press publishers are now entitled to restrain the unauthorized reproduction and making available of their online press publications by information society providers. Press publications are defined as a collection composed mainly of literary works of a journalistic nature, constituting an individual item within a periodical or regularly updated publication, providing the general public with information related to news or other topics and published in any media under the initiative, editorial responsibility and control of a service provider. Anyone wanting to use such a fragment with journalistic content, must obtain a license from the publisher of the publication in question.

The publishers’ right is limited to online uses of press publications and lasts for only two years after the publication date. Furthermore, this new right does not cover (i) private or non-commercial uses of press publications by individual users, (ii) acts of hyperlinking (and it remains unclear whether Google and similar services are technically doing anything more than hyperlinking while hosting the search engine website), or (iii) the use of individual words and very short extracts of a press publication. The notion of “short extracts” is not defined in the Copyright Directive and is likely to give rise to additional debate and a reference to the CJEU for interpretation.

This new right may not seem to be controversial at first sight, but many uncertainties remain to be clarified. Many researchers argue that it will not achieve its objectives, i.e., to ensure the sustainability of the publishing industry and foster the availability of reliable information. National experience in Spain and Germany, has already shown the likelihood for this right to backfire and further strengthen the power of media conglomerates and of global platforms (like Google) to the detriment of smaller players. Moreover, the new publishers’ right is said to encourage the use of fake news, as authors will be less likely to charge for the use of their publications.

3) Liability of user-upload platforms

Article 17 is probably the Copyright Directive’s longest provision, and it is certainly the most controversial. It has introduced new obligations for “online content sharing platforms” (OCSPs), which are defined as platforms with a profit-making purpose that store and give the public access to a large amount of works/subject matter uploaded by their users, which they organize and promote. This includes well-known platforms like Youtube and Dailymotion.

According to the new provision, OCSPs are liable for their users’ uploads, as it is stated that OCSPs carry out an act of communication to the public when they give access to copyright protected works and subject matter that is uploaded by their users.

In theory, OCSPs have two possibilities to avoid liability. First, they can try to obtain an authorization to communicate the content uploaded by their users and make it available. It is clear, however, that in practice it will be well-nigh impossible to obtain authorization for the millions of works that are uploaded. The second way to avoid liability would be for OCSPs to demonstrate that they have made “best efforts” to obtain permission from right holders, and to make unavailable and remove any infringing content when notified by rights holders, and keep the infringing content disabled.

Opponents have argued that this second possibility might result in censorship on the Internet. The OCSPs will deploy “upload filters”, which, by making use of automatic content recognition technologies, analyze uploaded content. These filters will ultimately result in general monitoring. Moreover, it is very questionable if such mechanism will be able to assess whether uploaded content is really infringing copyright, or rather falls under an exception (such as the parody exception).

More Partner Blogs


19 oktober 2020

Le “coronaouderschapsverlof” nouveau est arrivé: tijdelijke werkloosheid voor gebrek aan kinderopvang.

De regeling voor coronaouderschapsverlof is buiten toepassing getreden op 1 oktober 2020. Nu de...

Lees meer...

09 oktober 2020

The EU Copyright Directive: the three most controversial provisions

In April 2019, after years of debate and a rather controversial EU legislative process, the...

Lees meer...

24 september 2020

Can dismissals be justified based on Covid-19?

Can an employer dismiss an employee if the employee has tested positive for Covid-19 and refuses...

Lees meer...

17 september 2020

3 REDENEN OM E-MAIL TE MIJDEN

Waarom vergaderingen en e-mails niet samengaan. E-mail is een mooie uitvinding. Maar ook een...

Lees meer...

10 september 2020

DUALITY OF FUNCTIONS IN THE INSURANCE SECTOR: GAME OVER?

In this article we have a look at the consequences of a recent communication of the National Bank...

Lees meer...