Reflexos da inteligência artificial no direito penal: veículos autônomos e a responsabilidade criminal

• Ano de defesa: 2021
• Autor(a) principal: Vitor Eduardo Lacerda de Araújo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais
• Palavras-chave em Português: Direito penal; Responsabilidade penal; Inteligência artificial; Veículos autônomos
• Link de acesso: https://hdl.handle.net/1843/44791

Human beings call themselves homo sapiens because they understand that their intelligence is the differential for other living beings. For thousands of years, individuals have tried to understand the functioning of the human brain, that is, how a handful of organic matter, governed by laws of physics and chemistry, can perceive, understand nature, develop complex communication and language, predict events and manipulate a world much bigger and more complicated than its very essence. In the field of artificial intelligence, we find entities increasingly able to mimic the behavior of man, with significant evolution in diverse human competencies and cognitive abilities. In other words, with the sudden advance of computer science, man approaches the construction of machines endowed with rationality and capable of acting in a completely autonomous way, including attributes such as problem solving, perception, learning, creativity, action and reaction. Through data analysis and the consequent automation of the analytical models construction, as well as the identification of patterns, computerized systems are able to make decisions, such as the human being, with relative impossibility of predictability of the results arising from such conduct. Thus, considering machine learning based on human experiences, the problem arises about the consequences of artificial intelligence choices, that is, considering that human beings are constantly subject to crimes, there is a strong propensity for intelligent machines to commit them. In this sense, it can be mentioned that technologies recently introduced in human daily life, such as chatbots, unmanned aerial vehicles (UAVs) and autonomous cars, are able and already offend protected legal assets, such as honor, property and even life. Therefore, the study aims to investigate issues related to the causal relationships of the choices and actions of computerized artificial neural networks, especially those that constitute autonomous vehicles, aiming to investigate the consequences of the adoption of anti-legal conduct that result in criminal offenses caused by artificial intelligence, in the light of the theory of criminal responsibility.