Direito autoral das obras geradas por inteligências artificiais generativas: uma análise à luz da legislação de direitos autorais

Abstract

This research undertakes an analysis of the contemporary legal landscape of copyright and copyright, focusing on works produced by generative artificial intelligence (GIA). This work explores a constantly evolving problem: the difficulty of defining the author in works generated by IAG systems, such as ChatGPT1 (text generation) and Adobe Firefly2 (image generation). The topic in question emerges as a problem that challenges the traditional bases of copyright and intellectual property. As technology advances, the boundaries between human creativity and machine capabilities become increasingly blurred, posing a new challenge to legal systems around the world. In this context, the main question that arises is: who is the author of the works generated with AIG? The legislative uncertainty and the lack of legal precedents in this regard instigate debate among jurists and technology professionals, highlighting a significant gap in the legal system that requires urgent attention and possible reformulations, so that technological developments and their implications for the design of authorship and intellectual property. Based on the problem above, we intend to verify the hypothesis in the following sense: establishing an analogy between the user who interacts with IAG and the photographer who uses his camera to create a work, we seek to substantiate the attribution of copyright based on the concept of originality and creative contribution. In Brazilian copyright law, Law number 9,610/1998 provides that: Art. 7º Protected intellectual works are the creations of the spirit, expressed by any means or fixed on any support, tangible or intangible, known or that is invented in the future, such as: Thus, even if IAG carries out the materialization work, it is human intervention, the user's original and creative command, that defines the unique and authorial character of the work, that is, through the creation of spirit. Therefore, this perspective suggests that AI functions as a tool in the creative process, while authorial merit must fall to the individual who effectively conceived and orchestrated the creation, highlighting their creative expression and individuality in the command provided to artificial intelligence. This fact is due to the understanding that people only reproduce activities and do not reach the essence of the creative human being, as explained by Renata Lima and Marina de Almeida: In the case of the machine, Guattari (apud OLIVEIRA, 2005) understands that it tries to replace man's language not to emphasize what the subject is, but highlighting just what he does. In other words, at the moment when the computer does not seek the essence of the human being (or itself) and only repeats the activities that people do (imitation of actions), it can be said, in this way, that the sphere of classical ontology is repudiated (ALMEIDA and LIMA, 2020, p. 189). In this study, a hypothetical-deductive methodological approach is adopted to examine the hypothesis that the individual who enters the command into a generating AI (IAG) is the author of the works created, holder of moral and patrimonial rights, similarly to the photographer with his camera. Starting from the formulation of the hypothesis, logical deductions and verifications are carried out through comparative analyses, considering the interaction between operator and AI to establish an authorship relationship. Data collection will involve case law, legislation and case studies, analyzing authorship in AI-generated works and photography to find evidence that confirms or refutes our hypothesis, focusing on the interpretation of copyright laws. An analytical comparison is also carried out between the role of the AI ​​operator and that of the photographer, evaluating intention, creative control and contribution to the final result. Finally, a critical reflection on the findings will allow us to conclude on the validity of the analogy between the creator of the AI ​​prompt and the photographer, discussing practical and theoretical implications and contributing to the understanding of authorship in advanced technological contexts. This work will address the concept of copyright as defined in the Constitution of the Federative Republic of Brazil of 1988 and in Law number 9,610/1998, highlighting the protection of intellectual creations and the influence of the Berne Convention, in addition to discussing copyright following the perspective of José Carlos Costa Neto and analyzing international legislative differences. The definition of protected intellectual work and the originality criteria will be considered, according to article 7 of Brazilian law, and authorship based on article 11, exploring moral and patrimonial rights with references to João Henrique da Rocha Fragoso (2009) and Leonardo Machado Bridges. The study will include Alan Turing's (1950) vision of artificial intelligence, differentiating generative AI by diffusion and conversational, according to the thoughts of John Searle (1980) and will address the concept of machine learning according to Arthur Samuel. The use of IAG-based tools such as ChatGPT and Adobe Firefly will also be examined. in the context of copyright and artistic creation, concluding with an analysis of public domain and Creative Commons, referencing José de Oliveira Ascensão and Lawrence Lessig, to discuss their impact on the distribution of intellectual works in the digital era.