Artificial intelligence and robotics – a new regulation is coming

Artificial intelligence[1] (AI) is already part of our lives and adequate rules that foresees its rapid progression seems essential. 

The EU is rushing to create a regulatory framework that guarantees, in particular, the protection of consumers and citizens in general, the fight against unfair commercial practices and the protection of personal data and privacy[2]. However, these are not the only matters that will be subject to regulation and / or review by European organizations.

The EU Parliament has set up a special Committee on AI that approved on October 20, 2020, three reports on the regulation of artificial intelligence, about: (i) ethical respect, transparency and fundamental rights; (ii) civil liability and, (iii) industrial and intellectual property rights.

The Committee reminds us that the provisions of EU law regarding matters affected by AI will continue to apply, even if they require updates that reflect the digital transformation and use of AI in each field.

It is proposed to create a fully harmonized regulatory framework in the field of AI technologies that therefore takes the form of a regulation (non-directive) to avoid the fragmentation of the European digital single market.

The European Commission could present the first regulation on artificial intelligence next year.

Regarding the report on intellectual property rights, the importance of creating an effective system that allows the development of artificial intelligence to benefit the registration of Patents and new creative processes is emphasized.

Concerning to this report, we can highlight some issues raised by Parliament:

  • That the protection of Intellectual Property Rights must be reconciled with other fundamental rights and freedoms;
  • Recognizes the importance of the Patent system to encourage artificial intelligence inventions, and the role of essential Patents in the development and dissemination of novelties in AI and related technologies;
  • Underlines the legal challenges posed by reverse engineering, which is an exception to the protection of software copyrights and the protection of trade secrets, which are crucial for innovation and research.
  • Without prejudice to the provisions of the European Patent system, AI technologies must be freely accessible for educational and research purposes. 
  • Regarding who owns the intellectual property developed through AI, a differentiation will be made between human creations assisted by it and those generated by AI itself. The latter pose new regulatory challenges regarding the protection of IP rights (ownership, inventor status, adequate remuneration, and other issues related to possible market concentration). It is advanced that it does not seem appropriate to give AI technologies legal personality due to their negative impact on incentives for human creators;
  • Underlines that in those cases where AI is only used as a tool to assist an author in the creation process, the current IP Rights framework will continue being applicable;
  • Stresses the importance of facilitating access to and sharing of data, of open standards and open source technologies, and of encouraging investment and innovation at the same time;
  • Note that the AI used in application procedures for Intellectual Property Rights and in determining responsibility in situations of infringement of such IP rights cannot replace human review, carried out on a case-by-case basis, to guarantee the quality and equity of decisions.
  • Emphasizes that no AI regime can circumvent possible requirements related to open source technology in public tenders, or impede the interconnectivity of digital services;
  • Suggests, to the extent possible under Union law, that the largest possible volume of non-personal data be available for training and machine learning purposes and think about the use of public domain data for these purposes. Recalls for further clarification as regards the protection of data under copyright law, potential trademark and industrial design protection for works generated autonomously through AI applications.
  • It considers international cooperation on AI with organizations such as WIPO essential.
  • The objective is to achieve a balanced protection of IP Rights, based on innovation, that strengthens the international competitiveness of European companies, prevents abusive tactics in litigation, and guarantees maximum legal certainty. 

At the international level, WIPO, with the help of Member States, has compiled the main government instruments that have been produced on AI and IP. The Organization has created a specific space on its website that includes various resources available on AI in order to clarify the issue and facilitate the exchange of information between States.

How do creations made by or through AI affect IP Rights?

Systems based on AI could be divided into two main categories: (i) Software or computer programs (e.g. virtual assistants, image analysis programs, search engines and speech and face recognition systems); (ii) “Embodied” artificial intelligence, which is embodied in hardware devices (e.g. advanced robots, autonomous cars, drones or Internet of Things, etc.).

Regarding the protection of such creations, as regards embodied artificial intelligence (hardware devices), this may be patentable, in whole or in part, if they meet the requirements established by the legislation of each state. That technology that is not patentable, perhaps, could be protected by intellectual property rights, if the applicable regulations are complied with.

Concerning computer programs (software), although they are not protected a priori by means of Patent rights (they could be protected if it is shown that they meet the patentability requirements, such as solving a technical problem), also could be protected by intellectual property. 

What concerns to the authorship of the creations that artificial intelligence creates autonomously, beyond what they have been programmed for, it is debated whether the AI will have rights and how they can be materialized.

Considering the fast progress of the creations made by AI, which could go beyond what we currently imagine, the challenge will be to know where the machine created by man ends and where the autonomous entity and the creation it makes begins, and if these can be subject to rights in the future.

By now, the only possible scenario is that they belong to the creator or programmer who has created or designed them, or to the organizations on which they depend, the only ones which, at the moment, have legal personality and as understood, for the moment, by the EU Parliament.

We will see if recreations proposed by James Cameron in “Terminator”, or more contemporary ones, but no less disturbing, such as those proposed by “Black Mirror” (Netflix), predict an imminent reality. From a legal point of view, as lawyers we must understand the technology itself and be updated to provide an adequate added value.

___

[1] Artificial Intelligence, definition COM(2018) 237 final, p. 1: “Artificial intelligence (AI) refers to systems that display intelligent behaviour by analysing their environment and taking actions – with some degree of autonomy – to achieve specific goals. AI-based systems can be purely software-based, acting in the virtual world (e.g. voice assistants, image analysis software, search engines, speech and face recognition systems) or AI can be embedded in hardware devices (e.g. advanced robots, autonomous cars, drones or Internet of Things applications).”

[2] WHITE PAPER of the European Commission on Artificial Intelligence – A European approach to excellence and trust. Bruselas, 19.2.2020 COM (2020) 65 final.