Skip to main content


Report: CREATe researchers present on copyright exceptions for education and research at SCURL conference (Part 2)

Posted on    by CREATe Team

Report: CREATe researchers present on copyright exceptions for education and research at SCURL conference (Part 2)

By 10 July 2020March 18th, 2021No Comments

On 18th June 2020, CREATe researchers Bartolomeo Meletti and Thomas Margoni presented at the SCURL online copyright conference. The online event – organised by Greg Walters (University of Glasgow) and chaired by Jeanette Castle (University of the West of Scotland) – brought together around 50 information professionals from the Scottish Confederation of University & Research Library (SCURL). This is the second of two reports – the first can be found here.

Text and Data Mining – original illustration by Davide Bonazzi for

The subject of my presentation was the new EU Copyright in the Digital Single Market Directive (CDSMD) and its impact on post-Brexit UK researchers. After a brief overview of the relevant provisions of the Directive, I focused on Arts. 3 and 4 CDSMD, which introduce into EU copyright law two distinct Text and Data Mining (TDM) exceptions. The presentation also discussed how these articles will relate to Sec. 29A UK CDPA 1988 which, back in 2014, introduced an exemption for acts of non-commercial TDM in the UK. It was clarified at the outset that the provisions of the new CSDMD will most likely not be implemented in the UK in the foreseeable future. In fact, on the one hand, the UK will not be required to do so, since the term to implement the new EU rules (June 2021) is placed after the end of the transition period (December 2020), while on the other hand the UK Government has declared that there is no plan to voluntarily adopt those rules.

What will happen to UK copyright law after the end of the transition period, especially for cross-border uses so common in modern research collaborations of which TDM is the most iconic but certainly not the only example, remains an intriguing and uncertain question. The rule of thumb is that (most) of EU copyright law that is in force in the UK at the end of the transition date (31/12/2020) will be considered retained law and remain valid, either because it is directly applicable (EU Regulations) or because it has been implemented into UK law (EU Directives). Thus, not much will change immediately on 1 January 2021.

Given this situation, researchers active in the development of automated text and data analysis (not a small field) in the UK and in the EU will have to deal with the current UK and the upcoming EU TDM exceptions and on their cross-border (in?)applicability (see here for a recent paper discussing TDM and machine learning applied to legal research from a methodological point of view). These exceptions vary quite significantly in a number of aspects, especially in terms of beneficiaries and type of use: non-commercial research for the UK; research organisations for research purposes in the case of Art. 3, everyone for any purpose but with the option to opt-out in the case of Art. 4 in the EU. Another important element, which has been introduced into EU copyright law but is absent in the UK, is a provision that allows the storage of the training material for verification purposes. But there are also similarities between Sec. 29A and Arts. 3 and 4. For example the unenforceability of contractual provisions limiting the exceptions (obviously except for Art. 4), the ambiguous relationship to technological protection measures and perhaps more importantly the fact that all these provisions only exempt acts of reproduction, and not others such as distribution, communication to the public or adaptation (as discussed more extensively in this paper).  

Moving on to the broader (and more theoretical, but very important) issue of data ownership, which forms part of the work I am developing within the reCreating Europe project, I clarified how data as such is commonly excluded from copyright protection. International conventions are clear that ideas, procedures, methods of operation, mathematical concepts and the like are not protected by copyright. Only original expressions which constitute intellectual creations attract protection (see e.g. Arts. 2 WCT, 9(2) TRIPs, Art. 2 Berne and more recently Recital 9 CDSMD). Factual information and data as such commonly fall short of these requirements. Regarding data contained in databases, the EU SGDR (Sui Generis Database Right) protects original and non-original databases whose making required a substantial investment in obtaining, verifying, or presenting (but not in creating) the data. In this case, protection is against acts of extraction and reuse of substantial amounts of data. A third sub-category, information contained in a work, is usually not protected as such (logically belonging to the first category identified). Nevertheless, often an authorisation (e.g. exception, licence, etc) is necessary due to the fact that in the digital environment most uses require the making of a copy (even if temporary, indirect or in part) which is often an infringement of the copyright in the underlying work.

This explains the importance of exceptions and limitations such as the Text and Data Mining (TDM) exceptions. However it also highlights a paradox: information that is not afforded protection due to its crucial importance for fundamental rights such as freedom of expression and scientific enquiry becomes de facto protected in the digital environment when it is contained in qualifying works. Any future copyright reform should place the solution of this formal fallacy at its core.

In conclusion, I suggested that it is likely that while not much will happen on the first day of the new year, over the following months the UK and EU copyright systems will start to diverge significantly as EU Member States will implement the many changes brought by the CDSMD while the UK will not. It would be stimulating, albeit highly speculative, to imagine what changes UK copyright law will reserve for the future. Less speculative could be to imagine what forces will drive these changes: on the one hand the need to maintain a level playing field with the EU neighbour, a particularly important aspect for research and scientific collaborations especially during the ongoing health crisis. On the other hand the attractiveness of regulatory competition and the possibility to offer a more competitive and innovation-oriented copyright system in a post-Brexit UK. An example of the latter would certainly be the implementation of a broader TDM exception, not limited to non-commercial research (the current UK version), nor to research organisations for research purposes (Art. 3 CDSMD), or avoiding the fragmentation and uncertainty of opt-outs (Art. 4 CDSMD). Similarly, the exception should also not be limited to acts of reproduction, but cover all forms of reuse for TDM purposes. These will certainly be welcome changes for those who see copyright law as a balancing mechanisms between the protection of investments on the one hand and the promotion of innovation on the other. Only in this way can copyright law become what it once was, an instrument for economic, social and cultural growth.

The event concluded with a presentation by James Bennett (Head of Rights and Licensing at the Copyright Licensing Agency), who explained how some of the restrictions of the CLA Higher Education licence have been temporarily relaxed to provide support for the HE sector during the COVID-19 crisis.