Algorithmic discrimination as a form of structural discrimination: Standards of the Inter-American Court of Human Rights related to vulnerable groups and the challenges to judicial review related to structural injunctions

Abstract

In the era of the “information society”, many aspects of analogic life are migrating to the virtual space. Personal, educational, and professional relationships are becoming digital or hybrid. Although there are many conveniences in the virtual world, new dangers also arise, such as algorithmic discrimination. This occurs when the decision to grant or deny access to goods and services is made by an opaque or non-transparent algorithm. In this context, the following question arises: how can structural injunctions provide a basis for creating standards of protection in the face of legislative insufficiency and jurisdictional standards that confront algorithmic discrimination as a reproduction of structural discrimination? To this end, the hypothetical deductive approach, analytical procedure method, and bibliographic research technique were used. Thus, in the first part of this article, structural discrimination is contextualised, with special emphasis on the standards established by the Inter American Court of Human Rights. Next, the phenomenon of algorithmic discrimination and its relationship with historically observed structural discrimination in society are addressed. Finally, the possibility of using structural injunctions to provide Courts with an analysis of the phenomenon of algorithmic discrimination and formulate minimum standards to be observed by actors involved in the development of algorithms, especially those of artificial intelligence (AI), is examined. The conclusion is that structural injunctions adopted by Courts seeking to attack the roots of discrimination, especially those related to vulnerable groups, are an important tool to prevent algorithmic discrimination, as well as that the standards related to the protection of these groups established by the Inter-American Court of Human Rights should be taken as parameters
for legislation and regulation of the use of AI.