On 20 October 2020, the EU Parliament adopted three reports outlining how the EU can best regulate Artificial Intelligence (“AI”) while boosting innovation, ethical standards and trust in technology. The Parliament’s ambition is that of turning the EU into the leading worldwide’s player in the regulation of AI. One thing the EU appears quite good at is indeed the production of (much) regulation, not always paralleled by the EU businesses’ capacity of producing equivalent innovation. Whether there is correlation or a casual link between over-regulation and low innovation I do not know. If however one takes the example of personal data, it appears the case that the EU, while a world leader in the regulation of the sector, it is mostly a follower in the personal data driven economy. Leading the regulatory way might still earn the EU a pre-eminent role, not least because it provides a framework that other jurisdictions can replicate, as it happened with the GDPR, so that the EU practice and theories of the law may play-out to de facto harmonize the international legal terrain.
One of the three reports adopted by the Parliament concerns the interplay between Intellectual Property Rights (“IPRs“) and Artificial Intelligence. A topic I am very keen on, as is shown by my contributing a chapter on copyright and computer generated works to the Non-Conventional Copyright book published by Edward Elgar (https://www.e-elgar.com/shop/gbp/non-conventional-copyright-9781786434067.html). The Parliament gives interesting insights that, as legal intervention in the form of a regulation is called for, confirms the need to strike the appropriate balance between granting IP rights over the outcome of inventive or creative processes that have an AI content and preserving access to knoweldge and a healthy competitive environment.
The Parliament does provide some food for thought, which I can only distill here. Worth of note is in this connection the report’s preamble, where at letter F the EU legislator notes that AI and related technologies are heavily dependent on pre-existing content and large volumes of data. This is one of the main issues that I consider any future regulation of IP applied to AI should take into account. As a matter of fact, machines’ intelligence is premised on their capacity of mining and re-organizing massive volumes of existing knowledge/content, then the legislator’s concern should be to prevent any risk of indirect monopolization of public (information) goods. It is therefore aptly so that the Parliament notes how increased transparent and open access to certain nonpersonal data and databases in the Union, especially for SMEs and start-ups, as well as interoperability of data, which limits lock-in effects, will play a crucial role in advancing the development of European AI and supporting the competitiveness of European companies at the global level.
Linked to the above concern is the Parliament’s remark at paras. 14 and 15, where, with respect to coyrightable works, it indentifies a focal point: the need to differentiate “between AI-assisted human creations and AI-generated creations, with the latter creating new regulatory challenges for IPR protection, such as questions of ownership, inventorship and appropriate remuneration, as well as issues related to potential market concentration“. According to the Parliament, “where AI is used only as a tool to assist an author in the process of creation, the current IP framework remains applicable”. Yet the EU legislative body seems to recommend that technical creations generated by AI technology should be “protected under the IPR legal framework in order to encourage investment in this form of creation and improve legal certainty”. This calls for a reform of the legal framework in the EU, because, at least in certain Member States (and I believe this is the case for Italy), “works autonomously produced by artificial agents and robots might not be eligible for copyright protection, in order to observe the principle of originality, which is linked to a natural person, and since the concept of ‘intellectual creation’ addresses the author’s personality“; on this basis the Parliament “calls on the EU Commission to support a horizontal, evidence-based and technologically neutral approach to common, uniform copyright provisions applicable to AI-generated works in the Union, if it is considered that such works could be eligible for copyright protection”. The end of the sentence is somewhat disorentating, as it contrasts the initial encouragement to provide for a harmonzied regime of protection of works autonomously produced by artifical agent and ultimately appears to leave to the discretion of the Commission to determine whether they actually should be eligible for protection.
On my end, as I noted in the above mentioned publication, I believe that protection should be afforded to AI-generated works, to the person that has made the necessary arrangements (or even the user of the software), even though a lesser form of protection, such as a sui-generis right neighbouring copyright, shorter in time and based on the principle of protecting the investment, thus only allowing the enforcement of the right against parasitic forms of exploitation, rather than a full fledged right to control the acts that is typically bestowed upon copyright holders.