5 minute read 30 Apr 2020
dna research

Data Mining under the revised Swiss Copyright Act

By Andreas Jaeggi

Director, Head Digital Law | EY Switzerland

Experienced digital law expert. Focusing on ICT and privacy law and assisting clients in their digital transformation journey. Enjoys a global and multilingual environment, both at work and privately.

5 minute read 30 Apr 2020
Related topics Law

The amended Swiss Copyright Act now includes an exception for Text & Data Mining for scientific research, removing legal uncertainties.

In brief
  • The amendments to the Swiss Copyright Act came into force on 1 April 2020.
  • A newly introduced exception to copyright protection for scientific research includes text and data mining.
  • The exception permits technical acts of reproduction for research purposes and covers both commercial and non-commercial research. 

On 1 April 2020, the amendments to the Swiss Copyright Act (CopA), which the Federal Parliament had adopted last September after years of preparatory work, finally came into force. The primary aim of the revision was to adapt the CopA to the Internet age, for purposes of which compromise solutions were enacted in several different areas in order to address the divergent interests of authors, producers and users.

The present article deals with the newly introduced exception to copyright protection in the context of scientific research and its application to data mining and its subtype text mining, often referred to collectively as Text & Data Mining (TDM).

Fundamentals of Data Mining

TDM is an automated, computer-based method of analysing large amounts of text or data (big data) and interlinking them so as to discover cross-references, connections and patterns and thus to gain new insights. Whereas actual data mining is used for analysing large structured data sets, text mining is utilised on unstructured text files.

Because of the tremendous amounts of text and data available on the Internet, TDM has become a standard practice and is employed for all types of purposes, not only in scientific fields of research, such as the pharmaceutical industry and biomedicine, but also outside research, such as in marketing and journalism. Furthermore, TDM is used in developing applications based on artificial intelligence.

Relevance of copyright law

The TDM process normally starts with the creation of a data set on which the TDM software is then utilised for the purpose of analysing the data set. In this process, data are regularly copied and stored. If the data in question involve works as defined by copyright law (e.g. scientific or technical texts, images, audio-visual media or collections of data), the TDM process involves acts of reproduction of copyrighted works, e.g. through permanent storage of the works in the data set, through reformatting of the works contained in the data set or through their temporary storage during the analysis. Without either the author’s consent or statutory exceptions, such reproduction acts constitute copyright infringements.

Copyright regulation

Previous provisions in Switzerland

Prior to the revision, the CopA did not contain an exception to the author’s exclusive right in favour of scientific research or specific provisions on TDM.

Granted, acts of reproduction performed in the context of TDM and relevant to copyright law were sometimes privileged by the existing exceptions for personal use (for private and internal company purposes), as well as for temporary reproductions, and were thus permissible without the rightsholder’s consent. However, the consent of all affected rightsholders was required for all other uses of works or acts of reproduction for other purposes, such as the permanent storage of the data set. This always entailed a great deal of effort, high costs, difficulties in assessing the steps of the technical TDM process from a copyright standpoint and other legal uncertainties.

Regulation under the revised CopA

The principal goal of the comprehensive, newly introduced exception to copyright protection in the context of scientific research was to remove the aforementioned legal uncertainties existing in terms of the permissibility of TDM (and similar technologies) for research purposes under copyright law. The newly incorporated Article 24d CopA now expressly allows technical acts of reproductions (e.g. those performed via algorithms) for the purpose of scientific research, provided that the reproduced works have been lawfully accessed, as well as the storage of such reproductions for archiving and backup purposes once the research is complete. Consequently, such technical acts of reproduction are now lawful and royalty free even without the author’s consent. However, the reproduction of computer programs is excluded from this exception.

The new provisions cover both basic research and applied research. To avoid classification problems, the new statutory exception applies to both non-commercial and commercial research.

Regulation under the EU Directive

The statutory exceptions for the use of TDM were also put into concrete form at the EU level through Copyright Directive 2019/790, which came into force on 6 June 2019.

First, Art. 3 of the Directive allows TDM to be performed by research organisations that are non-profitmaking, re-invest all profits in their scientific research or operate in the public interest, as well as by cultural heritage institutions for research purposes, without the rightsholder’s consent. Second, Art. 4 allows TDM for any purpose, including purely commercial purposes, and thus goes beyond a mere exception for scientific purposes, but para. 3 limits the scope of these purposes such that the exception applies only if the respective rightsholder has not expressly prohibited TDM (reserved use).

Acknowledgement

Special thanks to Ramona Bollhalder for her valuable contribution to this article. 

Summary

On April 1, 2020, several amendments to the Swiss Copyright Act came into effect. One significant change concerns the new "research exception" provision, covering data mining and text mining for research purposes. These automated, computerized methods of analyzing large volumes of text or data create copies and store information, which represent an infringement of copyright law without the author's consent.

The updated act now makes these activities legal, providing

  • they are conducted for research purposes,
  • with legal access to the works,
  • and without infringement of any usage restrictions set by the copyright holder.

About this article

By Andreas Jaeggi

Director, Head Digital Law | EY Switzerland

Experienced digital law expert. Focusing on ICT and privacy law and assisting clients in their digital transformation journey. Enjoys a global and multilingual environment, both at work and privately.

Related topics Law