Responsabilidade penal de inteligências artificiais

• Main Author: Martins, Michael Menezes
• Publication Date: 2022
• Format: Bachelor thesis
• Language: por
• Source: Repositório Institucional da UFU
• Download full: https://repositorio.ufu.br/handle/123456789/35870
• topic: Inteligências artificiais; IA; Modelos de responsabilização penal; Machine learnig; Deep learnig; Teoria do crime; Culpabilidade; Artificial intelligences; Models of criminal responsibility; Culpability; CNPQ::CIENCIAS SOCIAIS APLICADAS::DIREITO::DIREITO PUBLICO::DIREITO PENAL

The technological evolution of recent years is remarkable. With robotics and nanotechnology studies, more and more limits, previously said to be unreachable, are being exceeded. Today the world is no longer the same as it was ten years ago, everything is changing very quickly and science is asking for passage. One of the most prominent fields is Artificial Intelligence. With a speech of breaking technological barriers and revolutionizing human relationships, the so-called AI's are still a nebulous topic for man. Great advancement has taken place from the earliest drafts of computer history to what we have today, with intelligent robots and algorithmic systems that act almost autonomously. Novelty and convenience are human vices, and having an intelligent being ready to serve human whims is nothing but “a dream”. For these reasons AI's started to enter people's lives, in their homes, in their jobs, in their car, in their social networks, there is practically some kind of AI in every modern computer program, of course more or less complex. That said, it appears that AI's are gradually becoming part of human daily life, and so to speak, building new legal relationships with them. Among the wide spectrum of law, one issue deserves to be highlighted when talking about AI’s: if an AI commits a crime, who should be held responsible? It is a difficult question to answer, but this work allows us to elaborate possible solutions to the problem, analyzing some models of criminal liability. It is also intended to reveal these enigmatic aspects of AI's, providing guidance on relevant topics such as machine learning and deep learning, and thus relate to the theory of crime and culpability in force in our penal code. . This work aims to start a very important debate for the present, which is already a social reality, and which will soon become an inevitable problem.