Generative artificial intelliegence challenges for intellectual property

Abstract

Generative artificial intelligence has come to the forefront of the international debate on artificial intelligence in general, on its challenges, risks and regulation. The availability of generative AI tools to society in general has highlighted some of the risks identified and warned about in recent years, especially those that may have an impact on copyright and fundamental rights, in particular the right to privacy, the right to information or the right to free artistic, literary, scientific and technical creation. The emergence of massively used generative intelligent systems based on large language models (LLM), deep learning and neural networks has shaken the basis of copyright as we know it and places us before a new scenario, in which any user, with or without specific knowledge and through a simple interface, can provide instructions (prompts) in text or audio form, of a technical nature or not, expressed in natural language and interpreted by them for processing and immediate creation of a result, which can range from a poem, a newspaper article, a synthetic image, a graphic design or a computer programme, to a comic, a song, a video game, a molecule, a new drug or a new lethal toxic agent. The international legal frameworks on intellectual property were conceived in a historical, technological, economic and social context that is very different from the current one, so that they cannot adequately respond to the different challenges and risks that generative AI may pose to it. Consequently, this paper aims to explore the main challenges that current generative AI poses in this area.