From Leonardo to the Next Rembrandt – The Need for AI-Pessimism in the Age of Algorithms

Abstract

Artificial Intelligence (AI) has become a part of our daily life, and “algorithmic creativity” has similarly gained a spotlight recently. From paintings to music, machines generate outputs that seem to comply with various prerequisites of copyright protection. The idea (and to some degree the need) to honour such achievements by legal protection has also emerged. This article demonstrates that contrary to the view of the proponents of an AI-copyright regime, time has not ripened for AI-copyright. The core elements of copyright law - namely the concept of authorship, originality and moral rights, as well as copyright’s history and incentives - are deeply rooted in an anthropocentric (although not only author-centric) world. Unless paradigm shift in copyright law, the lack of direct human element of an AI-generated output shall lead to the unavailability of copyright protection for these outputs. The article similarly highlights why the proposed alternative forms of copyright related rights or sui generis protection are doctrinally unfounded or practically unfit to provide for a sound solution that does not destroy the existing copyright regime and diminish the value of human creations.

The Influence of Artificial Intelligence on Copyright Law

“…In music, AI systems have composed pieces indistinguishable from those written by human composers, raising questions about originality and creativity (Duan et al, 2019). Visual arts have seen the emergence of AI painters that can produce artworks in the styles of historic masters, challenging traditional notions of artistic creativity (Mezei, 2020). Similarly, in writing, AI has been used to generate stories, poems, and news articles, some of which have been published by renowned media outlets.…”

“…While not a court case, it exemplifies the legal conundrums presented by AI in creative domains. The lack of clear legal guidelines on the copyrightability of AI-generated works leaves such matters open to interpretation and debate (Mezei, 2020). In the United States, the Copyright Office has issued guidance stating that works produced by AI without human authorship do not qualify for copyright protection, aligning with the traditional view that copyright requires human creativity.…”

Copyright in generative deep learning

“…The position of not assigning copyright in machine-generated works may appear to be convenient at first; indeed, it does not ask to change anything. It might also help preserve the centrality of human authorship in copyright law (see Mezei, 2020) and stress the importance of what an author should be versus what an author should do (see Craig and Kerr, 2019). Another, more practical reason is that a work should receive copyright protection only if an author exists; but to be considered so, the work must include a meaning or a message he/she wishes to convey, and this cannot happen if no one is able to predict the output of the program (Boyden, 2016), as in deep learning models (Ginsburg, 2018).…”

Reflections on the Use of Artificial Intelligence in Works of Art

“…In 2016, J. Walter Thompson Amsterdam and ING Bank collaborated to create a painting in the style of Rembrandt, using machine learning algorithms. The project began with the collection of data on Rembrandt's paintings, including his use of color, composition, and subject matter [16]. This data was then used to train a machine learning algorithm, which was designed to identify patterns in the data and generate new paintings that mimicked Rembrandt's style [17].…”