EU paving the way to regulate AI

Should AI be granted IP rights?

We are living in extraordinary times, times we might have dreamt of while watching Star Trek. No technological advancement has ever had an impact on everyday life to rival the revolution promised by Artificial Intelligence (AI). For that reason, it is essential that appropriate rules are adopted early on that, on the one hand, foster the creative exploitation of AI, and on the other, aim to preserve human rights and democratic values and thus encourage trust in it.

AI will shape the business of the future

AI is a major outgrowth of the digital transformation. Companies are already exploiting AI in chatbots that answer consumer questions, tools for predicting when a key sensor in a machine needs replacing and hence avoiding an expensive failure, and software that searches images to detect and diagnose cancer. The field of application appears unlimited. It is unsurprising, then, that many companies have already spotted AI’s considerable potential and – according to studies by Gartner and KPMG – have even increased their investment in the technology, despite the COVID-19 pandemic.

But companies often struggle to incorporate AI into their existing business models. On one hand, there are operational problems, such as a lack of knowledge of how to integrate AI into organisations and questions regarding financing. On the other hand, AI’s broad scope of application means that it will be utilized by highly regulated industries, thereby prompting a host of knotty legal questions.

Does the EU already know how to regulate AI?

The impact of AI was already so deep that the first intergovernmental policy guidelines on AI were adopted by the OECD in 2019. These guidelines provide international standards designed to ensure that AI systems are robust, safe, fair, trustworthy and respect our ethical values and applicable laws/regulations. These guidelines have already been adopted by Slovenia. Our country has recognised that as a multi-purpose technology, AI has the potential to improve people’s welfare and well-being, to contribute to sustainable global economic activity, to galvanise innovation and productivity, and to assist the response to key global challenges.

This October, the European Parliament adopted proposals that they believe will best regulate the field of AI to promote innovation in the European Union (EU), consolidate ethical standards, and strengthen confidence in technology. By applying these recommendations, the EU could become a world leader in the development of AI.

The first legislative initiative, brought by Iban García del Blanco, presents a new legal framework outlining the ethical principles and legal obligations to be followed when developing, deploying and using AI, robotics and related technologies (involving software, algorithms and data) in the EU. The core goal is that high-risk artificial-intelligence-based technologies – including any software, algorithms or data used or produced by such technologies – be designed, implemented and applied in a manner that guarantees full human oversight at any time.

The second legislative initiative, brought by Axel Voss, calls for a future-oriented civil liability framework, making those operating high-risk AI fully liable for any resulting damage.

The European Commission’s next step will be to prepare a legislative proposal that must consider the Parliament’s recommendations. Stéphane Séjourné’s report, however, touches an issue that may already have been partially solved by the European Patent Organisation (EPO).

AI and Intellectual Property issues

For the EU to become a true leader in AI, an effective intellectual property (IP) system working in parallel with appropriate protection measures is required to adequately protect both innovative developers and the EU's ethical principles. The particular significance of such balanced IP protection for AI technologies, and of its multidimensional nature, lies in the intention to secure future access to fundamental information, while simultaneously affording developers appropriate protection on the EU market. But including AI in a development process raises questions about the rightsholder.

Members of the European Parliament consider it important to distinguish between human creations supported by AI and creations fashioned by AI on its own. Members also stress that AI should not have a legal personality. Although the majority seem to believe only natural persons can own IP rights, this position has already been challenged.

Two appeals have already been filed against the EPO’s refusal to grant two patents in which the AI system “Dabus” is named as the inventor. The EPO rejected both applications on the grounds that the application does not satisfy the requirements of the European Patent Convention. Dabus has been rejected as an inventor in the UK, US and Germany on procedural grounds as well.

It is notable that IP5, that is the five offices that together handle most of the world’s patent applications, have set up a joint task force to explore the challenges posed by AI. Handling applications for inventions created by AI systems or machines is clearly high on their agenda. We may add that for now it seems unlikely that AI will be able to hold IP rights in the future. Certainly, such a change of view would fundamentally alter the development of future patent law and, as we all know, governments like to avoid setting such fundamental precedents.

We might be on the brink of new paradigm

Despite being far from implementing Isaac Asimov’s famous Three Laws of Robotics, we are unquestionably at the beginning of something new. With AI patents mushrooming and AI’s impact of on companies unmistakable, we will sooner or later need to tackle several key questions. One day the question might arise of whether AI-generated IP deserves its own sui generis law. The first steps towards the new paradigm will already be taken at the beginning of next year when the European Commission prepares its legislative proposal for regulation of the AI field, which we are all anxious to see as it will pave the way for the future development of AI. With the EU’s appetite for new tech legislation and its pressing determination to protect against algorithmic injustice, the EU wants to ensure we are ready for the challenges of AI as we enter our digital future.