The California court has made an initial ruling in the lawsuit filed by three American artists against the giants of generative artificial intelligence, Stability AI, DeviantArt, and Midjourney. The plaintiffs accuse them of having used their works to train the said AI programs and of allowing the generation of new works inspired by them through platforms that invoke their names.
Causes of action: copyright infringements and various violations
The artists base their action on different arguments to justify their case: direct and indirect copyright infringements, non-compliance with DMCA regulations, violation of their right to publicity, contractual breaches, and acts of unfair competition. Specifically regarding the copyright issue, the plaintiffs seek to demonstrate that generative AIs have indeed been trained with their works, relying on public statements made by the developers of these AIs.
The main issue: classifying AI productions as derivative works
The plaintiffs argue that each image produced by the generative artificial intelligence systems is exclusively derived from pre-existing latent images that are protected by copyright. Consequently, according to them, AI productions should be classified as derivative works requiring the consent of the authors of the original works.
The California court’s decision: a major turning point in the legal fight for the protection of artists
The California judge ruled on the motions to dismiss and strike filed by Stability AI Ltd., Stability AI, Inc., DeviantArt, Inc., and Midjourney, Inc. This initial decision represents a crucial step in this fight to protect artists’ rights in the face of technological advancements such as generative artificial intelligence.
A central point of conflict: the use of LAION databases to train generative AIs
Among the issues raised by the plaintiffs is the defendants’ use of LAION databases containing the works of the three artists to train their generative AI programs. This raises questions about the responsibility of the various parties involved – developers, users, and platforms that connect these two categories of actors – concerning compliance with copyright and intellectual property laws.
What potential impact on the generative artificial intelligence ecosystem?
As this legal battle is only just beginning, the stakes are enormous for the parties involved. On one hand, the plaintiffs seek to obtain compensation for their harm and to protect their rights as well as those of other artists in a similar situation. On the other hand, the giants of generative artificial intelligence will certainly be caught in a dilemma if this decision were to set a precedent.
The ethical question in the background: how far can we exploit works without infringing the rights of authors?
Beyond the purely legal aspect, this case highlights the ethical issue surrounding the {exp}loitation of artists’ works by machines equipped with artificial intelligence. How to find a balance between the necessity to innovate, particularly in the technological field, and the desire to protect and value the work of creators whose productions are often used without their permission or even their knowledge?
What does the future hold for copyright law and generative artificial intelligence?
This decision from the California court marks the beginning of a long legal fight whose outcome could have significant consequences for the development of generative artificial intelligence and the protection of artists’ rights. Whatever the results, it is clear that this case raises essential questions for the future of artistic creation in the face of technological advancements and their commercial exploitation.
