How will contemplated laws protect the results of AI activity in Europe? Considerations under selected provisions of EU law

Back to European Regional Forum publications

Aleksandra Auleytner
Domanski Zakrzewski Palinka, Warsaw
Aleksandra.Auleytner@dzp.pl

Marcin Jan Stepien
Domanski Zakrzewski Palinka, Warsaw
marcin.stepien@dzp.pl

Introduction

The legal protection of artificial intelligence (AI) in EU law is one of the European Union's current challenges. The rapid development of AI seems to require that certain regulations be adopted. On the one hand, this development within the EU should not be halted; on the other hand, it is necessary to ensure that the use of AI is safe and lawful. The use of AI raises the question about the rights to the results of its activity; that is, to the creations of AI. Most of these creations will be in the form of immaterial goods and their protection should be considered from the perspective of intellectual property law. These issues have already been recognised in the European Parliament resolution of 16 February 2017 (the ‘Resolution’) with the recommendations to the Commission on Civil Law Rules on Robotics.^[1] Nevertheless, this document does not in fact solve the actual question of the regulation of AI from the intellectual property law angle. In this context, several questions arise as to the direction in which the future regulation of the creations of AI should go and, specifically, whether this area should be protected under intellectual property law.

AI creations

Firstly, it needs to be explained what artificial intelligence is and what the nature of its creations are. In the Resolution of 16 February 2017, the European Parliament calls on the European Commission

‘to propose common Union definitions of cyber physical systems, autonomous systems, smart autonomous robots and their subcategories by taking into consideration the following characteristics of a smart robot: (-) the acquisition of autonomy through sensors and/or by exchanging data with its environment (inter-connectivity) and the trading and analysing of those data; (-) self-learning from experience and by interaction(optional criterion); (-) at least a minor physical support; (-) the adaptation of its behaviour and actions to the environment; (-) absence of life in the biological sense’.^[2]

In turn, in the Communication from the European Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee and the Committee of the Regions entitled the ‘Coordinated Plan on Artificial Intelligence’, it is stated that ‘Artificial Intelligence refers to systems that display intelligent behaviour by analysing their environment and taking action – with some degree of autonomy – to achieve specific goals. We are using AI on a daily basis, for example to block email spam or speak with digital assistants’.^[3] Thus, we may conclude that artificial intelligence is a computer system consisting of computer software, based on a source code, that displays intelligent behaviour by exchanging data with the environment and by analysing these data.

Data processing, as referred to above, constitutes the basis for AI action. As indicated in the literature:

‘Machines capable of automated inferences (AI) have a specific type of agency that can best be defined as data- and code-driven. They are data-driven since they can only perceive their environment in the form of data. Human beings perceive color, sound, contours, smells, tastes, touch, while our perception is always already mediated by language and interpretation. AI machines can only perceive any of this as data. This implies an act of translation or an environment that consists of data, for example, an online environment, virtual reality, or an IoT environment’.^[4]

This does not mean that the AI computer system cannot be embedded in hardware devices. For example, these can be devices such as smart robots, autonomous cars, or drones.^[5] Nevertheless, the natural environment of AI is still the virtual, digital environment.

Thus, most AI creations will also refer to the immaterial sphere. The protection of immaterial goods primarily relates to intellectual property law, including copyright law. In this context, it is no surprise that the European Parliament, in the Resolution of 16 February 2017, notes:

‘that there are no legal provisions that specifically apply to robotics, but that existing legal regimes and doctrines can be readily applied to robotics, although some aspects appear to call for specific consideration; calls on the Commission to support a horizontal and technologically neutral approach to intellectual property applicable to the various sectors in which robotics could be employed’.^[6]

However, provisions of intellectual property law, in particular copyright law, have historically been adopted to protect immaterial goods created by human intellectual activity.^[7] This assumption has also been confirmed by the author of the Berne Convention,^[8] which is an international act of fundamental importance for copyright law, including the copyright of the EU.^[9] The Court of Justice of the European Union ruled under the Berne Convention that ‘It is, moreover, apparent from the general scheme of the Berne Convention, in particular Article 2(5) and (8), that the protection of certain subject-matters as artistic or literary works presupposes that they are intellectual creations’.^[10] Additionally, in accordance with Article 1 section 3 of Directive 2009/24/EC of the European Parliament and of the Council of 23 April 2009 on the legal protection of computer programs: ‘A computer program shall be protected if it is original in the sense that it is the author's own intellectual creation. No other criteria shall be applied to determine its eligibility for protection’.^[11] As indicated in the literature, only a human is capable of creative activities.^[12] Consequently, it already seems quite difficult to protect the results of AI activity by copyright law.

Does this mean that AI products should, in principle, belong to the public domain? It seems that this should not be the case. As indicated in the literature: ‘Given AI productions costs are low, copyright duration can be shortened in the case of creative outputs. AI-generated works should not belong to the public, since the incentives to build the underlying AI-systems would disappear and end users would be reluctant to share their user-generated adaptations, gumming up innovation’.^[13] The creations generated with the help of computer programs are already present in today’s world. There are the creations such as paintings, newspaper articles, musical works, due diligence reports and financial reports.^[14] The laws of some countries already try to address this issue. For example, the UK Copyright, Designs and Patents Act 1988 protects computer-generated works. In accordance with Article 178 of that Act, computer-generated work ‘means that the work is generated by computer in circumstances such that there is no human author of the work’.^[15] These provisions of the UK’s copyright law are based, however, on specific provisions of the Copyright, Designs and Patents Act 1988, which are not present in the EU copyright law.

The recommendations from the European Parliament to the European Commission included in the Resolution of 16 February 2017 appear to confirm that the European law does not include legislation specifically applicable to robotics, including AI. It might be concluded that there are no specific laws concerning the results of AI activity, either. Further on, the European Parliament notes that ‘existing legal regimes and doctrines can be readily applied to robotics, although some aspects appear to call for specific consideration; calls on the Commission to support a horizontal and technologically neutral approach to intellectual property applicable to the various sectors in which robotics could be employed’.

This expressed intention of the European Parliament might lead to the following conclusion: in the absence of specific legal regulations that would govern artificial intelligence, we should reach for existing solutions (legal systems and doctrines) that can be applied accordingly to AI, including those which refer to the protection of immaterial goods, such as intellectual property law. Since it is accepted that AI regulations should remain in the realm of intellectual property law, the same approach could be adopted to the results of the AI activity. This means that should a complex regulation of the AI be adopted, the results of AI activity would need to remain in the scope of this regulation. It is therefore not advisable that a sui generis regulation of AI be adopted outside intellectual property law.

We remain of the opinion that the regulations contemplated to protect the results of AI activity may not be based on copyright protection. We believe that there is the risk that it would affect the paradigm for protection of human creations which already exists under copyright law. This is the critical fundament that we think may not be subject to modifications.

Considering the above, the following de lege ferenda conclusions could be formulated. The area of rights related to copyright seems the most appropriate one for the contemplated regulation. Related rights, being a part of intellectual property rights, have historically been conceived precisely as rights which exist ‘in addition’ to copyright, protecting immaterial goods similar to works (ie, those which are not exactly works in the meaning of copyright law but which, because of their particular situations, require protection under exclusive rights). For instance, regulations on related rights relating to phonograms confer exclusive rights on producers of the phonograms, even though they are not the authors of the phonograms, but because they spent their time, effort or financial outlay for the production of the phonograms.^[16]

We also note that the related-rights approach is also beneficial in that the provisions on related rights, in principle, do not exclude the simultaneous legal protection of the results in question on different legal bases should those results or their elements be suitable for such protection, for example, under copyright law. This may turn out to be important in the case of the creations that arise as a result of the activity of both AI and a human being, which, due to certain provisions on copyright, could also be protected under this area of the EU law.