Book review: Intellectual Property Protection for AI-Generated Creations

Having recently reviewed The Future of Copyright in the Age of Artificial Intelligence, here, this Kat continues thematic reviews in this area with a review of Intellectual Property Protection for AI-Generated Creations: Europe, United States, Australia and Japan, by Ana Ramalho (Copyright Counsel at Google, and a Guest Lecturer at Leiden University, The Netherlands). 

 

Ramalho’s structure is simple and clear, two main parts, in two chapters – the first covering AI and copyright, and the second AI and patents – with one agenda, are AI-generated creations protectable? 

Both chapters follow the same recipe. Each considers the way in which each AI can generate products that, if created by humans, would qualify for protection under copyright or patent. Ramalho then examines the justifications for granting copyright or patent protection with a view to considering  whether that protection can be extended to include AI-generated content. 

She then analyses relevant parts of copyright and patent protection for AI-generated content under European, US, Australian and Japanese law. For the copyright chapter, the relevant elements included subject matter, originality, authorship and neighbouring rights. In this chapter, Ramalho recommends that whilst copyright should not protect authorless works, an eventual neighbouring right might be enough to balance the interests at stake.

The patent chapter addresses patentability subject matter, inventive step or non-obviousness, enablement or sufficient disclosure, inventorship, and other forms of protection such as utility model. In this chapter, Ramalho explains that the questions around AI and patents are different to those relating to copyright, because in patent law a human is needed to formulate the problem, which is a relevant stage of the inventing process. Likewise, it is also a relevant stage of the inventing process where a human is needed in machine learning, because, she states, the process of getting a patent implies a prior art search. 

Her analysis of the different approaches taken across the four jurisdictions leads Ramalho to conclude that “there is a relatively high level of international harmonisation of current patent law.” She goes on to argue that whilst AI systems that generate inventions might not be completely autonomous, “it is undeniable that they augment the inventive capabilities of human beings.” The chapter concludes with recommendations such as that Patent Offices should revise and update their guidelines, specifically concerning inventive step and non-obviousness. 

In the overall conclusion, Ramalho draws together her key arguments and common themes from the two main chapters. She also highlights areas in need of further investigation, such as ownership and moral rights. 

An important pattern that cuts across patent and copyright emerges: the creation of new rights as an option for protection of new realities remains always a possibility, but it is not something that should be taken lightly.

This book will therefore appeal to researchers and policymakers interested in the intersection between artificial intelligence and intellectual property rights. 

Routledge

Available in hardback and eBook 

ISBN 9780367415617

Extent 172 Pages