Copyright and Artificial Intelligence: between legal frameworks and innovation

Copyright and Artificial Intelligence: between legal frameworks and innovation

Generative Artificial Intelligence is a major step forward in terms of new technologies, raising concerns over our current legal frameworks. And as lawsuits are being filed against AI giants (i.e. OpenAI), Pleias, a French startup, innovated in the creation of generative AI to only use unprotected data to respect copyright. Furthermore, the French National Assembly’s Higher Committee on Digital and Posts has published proposals for revising legislation such as the European Copyright Directive, to be more inclusive of AI.


Generative Artificial Intelligence is now making its mark in almost all fields. Indeed, it is capable of creating songs or images, as well as automatically generating well-written texts from data it was trained on. Consequently, the distinction between human creation and creation by AI is becoming increasingly blurred, raising legitimate concerns about security and transparency. But also, about intellectual property rights and, more particularly, copyright.

This is happening especially because giants of the AI industry like OpenAI, Anthropic are currently claiming that it is impossible not to use data subject to copyright and, as a result, many of them are facing lawsuits, mostly in the United States, about copyright infringement (i.e. “The New York Times Company v. Microsoft Corp., OpenAI Inc. et al., file on the 27th of December 2023 with the New York Court[1]).

Meanwhile, to avoid this risk, a French startup named Pleias, has developed an interesting project:

an open Common Corpus text that can train generative AI to only use data that has fallen into the public domain in order to ensure the respect of copyright[2].

Pierre-Carl Langlais, Anastasia Stasenko and Prof. Dr. Ivan Yamshchikov, the cofounders of Pleias, supported by many research partners[3], launched an innovative and ethical corpus on March 26, 2024, which contains about 500 billion words from several languages like French, English and Italian, and over 300 billion words from French and European administrative and legislative texts.

Although this is encouraging, it doesn’t take away the need for a more precise and clearer legal framework for the use of AI. In this context, the French National Assembly’s Higher Committee on Digital and Posts (CNSP) have published 30 recommendations in its n°2024-01 Opinion, on January 17, 2024[4].

One of the recommendations made by the French CNSP / the French National Assembly’s Higher Committee on Digital and Posts precisely concerning “the regulation and governance of artificial intelligence systems” suggests revising the European Copyright Directive in order to include generative AI, in order to take into account the great impact of generative AI on copyright and, more generally, on intellectual and industrial property. Members of the French Parliament therefore consider that it is essential to envisage a revision of the Directive during the next European legislative term, as it would fill a legal vacuum.

They also believe it would be appropriate to consider a frequent review of the forthcoming European AI Act[5] to include all AI’s evolution. As it would continue to encourage research and innovation in the field of AI while ensuring an effective protection of copyright. And, on a more general level, a protection of intellectual and industrial property.

Setting up a long-term robust yet flexible legislative framework would make it possible, not only to adapt the current legislation, but also to anticipate the future challenges posed by the fast evolution of AI.


Di Maiwenn Gervais
French student in a double degree in Law and Applied Foreign Languages at the University of Brest, and current legal intern at ALLegal law firm in Milan, Italy


Note:
[1] M.M. Grynbaum – R. Mac, New York Times Sues OpenAI and Microsoft Over Use of Copyrighted Work, in New York Times, 27th December 2023.
[2] F. Niedercorn, Pour une IA générative qui respecte le droit d’auteur | Les Echos, in Les Echos, 26th March 2024.
[3] Information from the website of the startup, Pleias, founded by Pierre-Carl Langlais, Anastasia Stasenko and Prof. Dr. Ivan Yamshchikov.
[4] Commission Supérieure du Numérique et des Postes (CSNP), Avis n° 2024-01, Pour mieux encadrer l’usage de l’intelligence artificielle, 17th January 2024 (Avis – CSNP).
[5] European Parliament, Artificial Intelligence Act. European Parliament legislative resolution of 13 March on the proposal for a regulation of the European Parliament and of the Council on laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union Legislative Act, 13th March 2024.