FEB 06, 2019 | EUROPEAN UNION
BSA urges EU Institutions to Enact Mandatory Exception for Text and Data Mining In Copyright Directive
BRUSSELS - February 6, 2019 –Ahead of the upcoming 8 February COREPER meeting, where Member States will vote on a renewed mandate for negotiation on the Copyright Directive, BSA | The Software Alliance urges all three EU Institutions to enact a mandatory exception for Text and Data Mining (TDM) for commercial purposes in Art. 3(a) of the Directive.
TDM is a form of software-enabled analytics that enables unlocking correlations and identifying useful knowledge from information that rests undiscovered in data sets large and small, in ways humans can process and harness for a myriad of valuable purposes. TDM is one of the revolutionary tools used by Artificial Intelligence, a fundamental process in machine-learning and data analytics.
The EU Institutions are currently debating on whether to enact a mandatory exception for Text and Data Mining. As a global organization, BSA has advocated for broad TDM exceptions (only in cases where the party doing so already has lawful access to the underlying work) with all global leaders in AI, with countries such as Japan, Singapore, China and the United States currently providing broad exceptions. At the same time, Canada and Australia are considering updates to their Copyright legislation, including an exception for TDM for commercial uses. With all global leaders moving towards balanced legislation in this field, the EU has the opportunity to claim a leading role in AI development globally.
“With the Trilogues on the Copyright Directive coming to an end, the EU finds itself in the position of deciding whether or not to take a leading role in Artificial Intelligence, creating future-proof rules that protect rightsholders while supporting cutting-edge technologies and research” said Thomas Boué, BSA Director General, Policy – EMEA.“ We urge the EU Institutions to strongly support a mandatory exception for Text and Data Mining in Art. 3a, and ensure that EU-based Industry and Researchers are given the necessary tools to compete globally.”
The most recent development in this field was in Singapore, in January 2019. Singapore’s Ministry of Law has on January 17, issued a report containing its policy decisions on its 2016 public consultation on proposed changes to Singapore’s Copyright Act.
With regards to Text and Data Mining, the Ministry has decided to adopt a broad exception for TDM for data analysis, covering both non-commercial and commercial non-consumptive use of copyright works.
In May 2018 Japan passed an amendment to their Copyright Act, creating various broad and new exceptions aimed at facilitating technological uses of copyrighted works, including TDM. Specifically, it now provides for an exception for non-consumptive uses of copyright works – i.e., uses “where it is not intended to enjoy or to make others enjoy thoughts or sentiments expressed in a work.” The provision points to three specific uses cases that would qualify for the exception, including “where the work is used for information analysis (meaning the extraction, comparison, classification, or other statistical analysis of language, sound, or image, or other elements of which a large number of works or a large volume of information is composed).” This covers the full range of activities involved in machine learning process (including TDM), and does not distinguish between non-commercial or commercial applications of the results of such activities.
The text of the as-passed legislation is available here. For more information, please refer to our TechPost on the new developments in Japan, and how they put Europe at a disadvantage in the original formulation of the Copyright Directive, which can be found here.
Canada and Australia are in the process of reviewing their Copyright rules, and are considering enacting broad exception for commercial and non-commercial uses of TDM. While still behind in the process compared to the EU, both countries have shown the intention to support AI development and strengthen their position on global markets.
Lastly, the United States currently provides for a broad exception for Text and Data Mining under the Fair Use Doctrine. This reinforces the global primacy for the United States in Artificial Intelligence, as most global leaders in Artificial Intelligence are either based there, or in the process of moving significant portions of their AI development operations there. Similarly, China provides for broad exceptions for commercial and non-commercial uses for TDM.