EU Copyright Directive – A statement of the student council of Computer Science

Over the last two years, the European Parliament has been working on a reform of copyright law for the digital age. The final draft was adopted in the trialogue on 13 February and will be voted on by the European Parliament on 28 March.

The student council of computer science at the TU Kaiserslautern welcomes the idea of adapting copyright law to today’s circumstances in which everyone can be a producer and author on the Internet. However, the proposed Directives force us to oppose the Copyright Directive. Two paragraphs in particular worry us: Article 11 (protection of press releases concerning digital use) and Article 13 (use of protected content by information society service providers [...]).

Article 11 allows press publishers to charge money already for displaying extracts from their articles. Based on our experience in Spain and Germany with similar laws, we consider this to be ineffective. In addition, we see a danger for small publishers who, unlike large competitors, are disproportionately dependent on users coming via search engines.

Article 13 requires that any works uploaded by users of Internet platforms (texts, images, videos, etc.) may only be published if appropriate exploitation licences have been negotiated with the rights holders of the works. Like the United Nations High Commissioner for Human Rights, we regard this as a danger to the freedom of art and speech [1]; thus platform operators such as YouTube and Facebook must introduce censoring upload filters in order to comply with the regulations in Article 13. However, these are generally associated with high error rates and cannot handle quotes and parodies.

As computer science students, we know which problems can be solved with the help of statistics, machine learning and artificial intelligence – ultimately with the help of algorithms. In the case of Article 13, algorithms must decide whether the licences mentioned above have been negotiated. This is typically achieved by recognizing examples, i.e., the algorithm already knows a catalogue of works with a negotiated license agreement and recognizes whether the newly uploaded work corresponds to one of the previous ones. However, these algorithms can only ever find such a match with a certain degree of certainty, i.e. with a certain probability they make mistakes. This means that many legitimate publications would automatically be censored, although there is no reason to do so. Furthermore, the Commission promises that art and satire and the ability to quote will not be compromised by this reform. Even if an algorithm could make perfect licensing decisions (which is not foreseeable in the medium term), it would still have to be decided whether a new work should be used in the context of artistic freedom and not prematurely deleted.

In addition, we agree with the Federal Data Protection Commissioner that the implementation of Article 13 in practice will mean a further centralization of power and information with large Internet corporations, which in the future will also judge content that has not been uploaded to their platforms [2].

In view of the far-reaching consequences of the Copyright Directive, we call on the Members of the European Parliament to vote against this Directive on 28 March.

Literature:
[1] David Kaye, EU must align copyright reform with international human rights standards, says expert, https://www.ohchr.org/en/NewsEvents/Pages/DisplayNews.aspx?NewsID=24298.
[2] Der Bundesbeauftragte fĂĽr den Datenschutz und die Informationsfreiheit, Reform des Urheberrechts birgt auch datenschutzrechtliche Risiken, https://www.bfdi.bund.de/DE/Infothek/Pressemitteilungen/2019/10_Uploadfilter.html.

Translated with www.DeepL.com/Translator