Edition: 187 | Month: November | December | Year: 2023
By Laetitia d’Hanens and Maitê Cecilia Fabbri Moro
Generative AI: Techniques, Opportunities, and the Challenges of Authorship
By Diogo Cortiz, Jefferson Oliveira and Kizzy Terra
The paper explores recent techniques in Generative Artificial Intelligence (GAI), emphasizing their significant advancements and implications. After introducing the concept of Artificial Intelligence, the text traverses the history of AI and highlights innovative models driving GAI, including Foundational Models, Language Models, Generative Adversarial Networks, and Diffusion Models. The transformative potential of GAI across various societal spheres is underscored, addressing both the opportunities and risks associated with these changes. Finally, the article provides insights into the discussions around authorship and proposes reflections on the use of AI from an interdisciplinary perspective.
Keywords: Generative Artificial Intelligence. Foundation Models. Diffusion Models. ChatGPT.
Text and data mining and copyrighht
By Manoel J. Pereira dos Santos
This article analyzes the impact of Text and Data Mining in copyright, discussing the situation of the matter under current law, in Brazil and abroad, and the proposals aimed at creating legislation for a specific legal permission so that carrying out these activities is not considered a copyright infringement.
Keywords: Copyright. Author’s rights. Artificial Intelligence. Text mining. Data mining. TDM. Data bases. Big Data.
Generative artificial intelligence and challenges for intellectual property: copyright, regulation and legal protection
By Juliana Abrusio and André Gualtieri
Generative artificial intelligence has raised questions about central concepts of copyright legislation, such as authorship and originality. This article examines the challenges brought by generative AI for the legal protection of intellectual property. Firstly, it discusses problems related to authorship and ownership of rights in works created by AI, as well as to infringement of rights both in the training process of these systems and in their outputs. Secondly, it compares how different legal systems have faced these issues, drawing an overview of the regulatory approaches on AI in the European, North American and Chinese contexts regarding the impacts on the intellectual property area. It concludes that the complexity of AI systems requires a broad understanding of the topic, combining technical, ethical and legal aspects. It advocates the need for appropriate regulation, capable of ensuring a fair and balanced protection in the face of the challenges and transformations promoted by technology.
Keywords: Generative Artificial Intelligence. Copyright. Intellectual Property. Authorship. Regulation.
Generative Artificial Intelligence and Possible Patent Infringements: A Study in the Brazilian Context
By Tarso Mesquita Machado and Daniel da Silva Costa
This paper delves into the intersection between Generative Artificial Intelligence (GAI) and patents in Brazil. Recent discussions on GAI and copyright are introduced, followed by an exploration of challenges related to patents, particularly issues concerning patentability, authorship, and the possibility of infringement. Three exemplary scenarios are outlined and evaluated with a practical approach under the Brazilian Industrial Property Law (LPI). It is emphasized that the mere creation of an invention by GAI is not capable of infringing a patent in Brazil, as the act of creation or publication does not fall within infringement possibilities of our Law. Still, the use or exploitation of the invention created by GAI without considering preexisting patents may lead to patent violations, which raises awareness for the importance of care when using such innovations and the importance of carrying out freedom to operate analysis. Therefore, like other independent creations, an invention created by GAI does not have presumption of patentability or that there is freedom to use the technology in the country.
Keywords: Generative Artificial Intelligence. Patent Infringement. Machine Learning. Neural Networks. Intellectual Property.
Personality rights and civil liability in the use of artificial intelligence
By Filipe Fonteles Cabral
The article discusses civil liability arising from damages caused by Artificial Intelligence (AI) systems, with special attention to situations of violation of personal rights. The study presents the current state and types of AI under development. Next, it addresses the personal rights potentially impacted by the use of AI, the European regulatory initiative and doctrine reflections on new regulatory models. The work is concluded with an analysis of the application of civil liability for damages caused by AI under the Brazilian legal system, including comments on Bill #2,338/2023.
Keywords: Artificial intelligence. Civil liability. Personality rights. Regulation.
Trade secret and algorithmic transparency in view of the Brazilian Data Protection Law (LGPD)
By José Eduardo de V. Pieri, Fabio Pereira and Vanessa Ribeiro
The trade secret safeguards the confidentiality of valuable information kept confidential to protect the company’s competitive advantage and that has commercial or industrial value. There is an undeniable intersection between the LGPD and the trade secret that involves personal data, intensified in recent years by the use of artificial intelligence systems and the consequent need for algorithmic transparency. The article, therefore, aims to contribute to the debate on the mentioned intersection, providing guidance on the possible reconciliation of the mentioned presumed conflicting interests, particularly in light of the provisions of the LGPD.
Keywords: Trade secret. Intellectual Property. Unfair competition. Personal data. Inferences.
Challenges in the ABPI’s 3rd Moot on addressing issues concerning Artificial
Intelligence in Intellectual Property
By Isadora Staeheli, Gabriel Milton Parente Araújo, Rodrigo Ninomiya Boin, Milena Laranjeira Vilas Boas, Letícia Thiemy Venturini, Natália Moraes Ferreira, Eduarda Salema Raposo, Ana Laura de Souza Pereira Pacheco and Cassius Jansen Coelho da Silva
This article addresses the challenges faced by the winner team of students during the 3rd ABPI Moot in discussing complex issues related to artificial intelligence (AI) in intellectual property. The hypothetical case involves the paternity and ownership of an invention generated by an AI named DAIA. The team debates the relationship between the creator, Dr. Stefano, and the AI, issues regarding the autonomy of the AI in creation, remuneration of the creator as both a partner and inventor, the legal recognition of an AI as an inventor, and the need for agile regulation in the face of fast AI advancement. The article also highlights the team’s preparation for the Moot, including the questions that generated the most uncertainty among students, what they did to develop convincing arguments on controversial issues and lingering questions after the competition’s conclusion.
Keywords: Artificial Intelligence. Intellectual Property. Inventor’s Rights. Paternity. Moot Court.