O direito à revisão humana em decisões automatizadas: o contorno jurídico da LGPD e as possíveis consequências ao titular de dados

Abstract

This article aims to address the existing problems surrounding artificial intelligence technologies and machine learning for data subjects exposed to the possible and resulting discrimination of algorithms in automated decisions as well as damage to selfdetermination and personality rights. As a legal instrument to protect data subjects, there should be the right to human review in order to minimize damages; however, this provision was removed from the General Law of Data Protection, sanctioned in 2018. Based on this, this paper intends to expose, the questions regarding this science and the criteria present in the technology, as well as the possible defects and damages to the holders arising from automated decisions. To do so, the deductive methodology was used, based on an analysis of the legislation, as well as on bibliographic and doctrinal consultations that discuss the theme, in order to substantiate the adopted position. As a result, the intense discrimination and technical flaws that can result from technology were identified, especially without the legislative protection in providing the right to human review in automated decisions.