ARTIFICIAL INTELLIGENCE AND LABOR

• Autor(a) principal: Fornasier, Mateus de Oliveira
• Data de Publicação: 2021
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Revista da Faculdade Mineira de Direito (Online)
• Texto Completo: http://periodicos.pucminas.br/index.php/Direito/article/view/23491
• topic: artificial intelligence; social security; universal basic income; discrimination; transparency; inteligencia artificial; seguridad social; renta básica universal; discriminación; transparencia; inteligência artificial; seguridade social; renda básica universal; discriminação; transparência

This article studies the impacts that the new economy, resulting from automation, represents for Law. Its hypothesis is that a new social security, based on universal basic income, funded by taxing the use of automation tools, should replace systems based mainly on the employment relationship; and principles related to transparency, explicability and non-discrimination should create obligations for developers and users of AI-powered worker selection tools. Methodology: hypothetical-deductive procedure method, with a transdisciplinary and qualitative approach, and bibliographic review research technique. Results: i) Labor regulation must be planned beyond substitution, focusing on a new economy, in which formal jobs, inserted in a paradigm of social and economic protection, are eroding, and the great challenge will be to protect decent work standards concurrently with the enlargement of dignity for non employed workers; ii) A universal basic income funded by taxes on automation would be interesting, but problematic from the point of view of solidarity in its costing, since the stress of national economies can cause fear on governments. And globally, such taxation, if adopted differently, can cause great tax competition between countries; iii) Transparency obligations are necessary to mitigate bias in hiring tools, but not self-sufficient, because the bias is complex, mainly due to the multiplicity of discriminatory factors and to the opacity of the logic in machine learning.