Artificial Intelligence and Copyright
Artificial Intelligence (AI) represents one of the most significant technological transformations of our time. Its entry into everyday life has made possible activities ranging from automatic translation to the production of literary, musical, visual, and cinematographic works. However, this generative capacity raises legal issues, particularly concerning the ownership of copyright in works so produced and the compliance of the datasets used to train the models.
Generative AI systems operate through deep learning, a process that analyzes enormous amounts of data and creates connections among them, thereby generating new content. The centrality of data raises an initial legal issue: if such datasets include protected works, their use without authorization entails the risk of copyright infringement. The distinction between mere “training” and creative exploitation thus remains one of the most delicate aspects, and one that continues to be debated in courtrooms.
The European Union has partially addressed the matter with the AI Act, approved by the European Parliament on March 13, 2024, and entering into force in 2026. The regulation requires AI providers to comply with EU copyright law, adopt appropriate policies, and ensure that the datasets used are lawful. This strengthens protection for authors, preventing AI-generated works from being based on material obtained without consent.
A central unresolved issue, however, remains the authorship of the work. Current regulations and laws, drafted with human-created works in mind, do not provide exhaustive answers. Several hypotheses have been put forward: attribution of authorship to the AI (excluded, since it is not a natural person), to the developer, to the database owner, or to the user who inputs the prompts. The most widely accepted solution today is to recognize protection only for works in which the human contribution is creative, relevant, and demonstrable. In Italy, the Supreme Court (Cass. civ., Sez. I, Ordinance, 16 January 2023, no. 1107) clarified that the use of software in the creative process does not in itself exclude copyright protection, provided that the human creative contribution remains predominant over the role of AI systems. A similar approach has emerged in U.S. law, where the case of Zarya of the Dawn set an important precedent.
The problem of authorship is closely connected to the notion of creativity, which encompasses freedom of choice and discretion on the part of the human author. AI, by operating on pre-existing data, risks undermining the requirement of originality. Nevertheless, it must be remembered that even human artists are influenced by their own cultural and artistic background. For this reason, European and Italian legislators have suggested considering AI as a tool in the creator’s hands, comparable—at least to some extent—to the use of a camera.
The use of AI also entails circumstances that may be regarded as risky, including voice cloning and the reproduction of faces, which raise serious concerns about the use of an individual’s image and personal data when digitally replicated. The film industry provides an emblematic example of the impact of AI. The arrival of AI has enabled the writing of screenplays, the creation of visual effects, and even the digital replication of actors and dubbing artists. The reaction of the industry has taken the form of strikes and the conclusion of union agreements in the United States, introducing rules on digital replicas and vocal alterations, with the consent of the artist placed at the center of the regulation.
Technological evolution has always generated fears and resistance. In the 19th century, John Philip Sousa referred to phonographs as “infernal machines,” convinced that they would destroy music. History proved him wrong: technology contributed to the diffusion and survival of musical art. In the same way, AI can be regarded as a new creative tool, capable of supporting the author without replacing them. The legislator’s task is twofold: to safeguard established copyright and to regulate new forms of digital creativity. Within this balance lies the most urgent challenge of contemporary law: reconciling the protection of the individual and their work with the opportunities opened up by technological innovation.
