Inteligência artificial no recrutamento e o risco de discriminação algorítmica: limites jurídicos e responsabilidade dos empregadores à luz dos casos Amazon e Kyle Behm

Abstract

With the advent and popularization of artificial intelligence, the use of algorithms for resume screening and automated decision-making in recruitment has become increasingly common. Despite the technological advances, challenges arise regarding the risk of algorithmic discrimination. In this context, the present research investigates how the automation of recruitment processes through artificial intelligence can result in discriminatory practices and what legal limits exist for employer liability due to algorithmic discrimination under Brazilian law. The study is based on the premise that algorithms can reproduce structural discriminatory practices, even unintentionally, due to biases in training data. Methodologically, it is a qualitative, basic, and exploratory research, conducted through a bibliographic procedure with analysis of books, scientific articles, legislation, and national and international jurisprudence, as well as the study of emblematic cases such as Amazon and Kyle Behm. The results show that, despite the lack of specific legislation on artificial intelligence in Brazil, employers can be held accountable based on the Federal Constitution, the Civil Code, the CLT, and the General

Data Protection Law, especially considering the principles of human dignity, equality, non- discrimination, and transparency. It is concluded that employer liability may be configured

under both subjective and objective modalities, the latter being applicable when the use of AI in selection processes is considered a risk activity, thus requiring greater diligence in choosing, implementing, and monitoring such technologies to prevent algorithmic discrimination.