Racial perspective in the dialogue between the Declaration of Principles on human rights in the field of artificial intelligence and the brazilian constitution

Abstract

This article examines the dialogue between the Declaration on Human Rights Principles in the Field of Artificial Intelligence (DPDHAIA) of Mercosur, promulgated in 2023, and the 1988 Federal Constitution of Brazil, focusing on the protection of the fundamental rights of the Black population in the face of AI-related impacts. The study investigates the extent to which the principles of the DPDHAIA find support in Brazilian constitutional provisions, considering the vulnerabilities of racialized populations. The objective is to identify correspondences and normative gaps in addressing structural racism mediated by digital technologies. This is a qualitative, exploratory, and documentary research. The methodological approach consisted of analyzing the full texts of the DPDHAIA and the 1988 Constitution, guided by the search for keywords (“race,” “dignity,” “equality,” “racism,” “prejudice,” and “human rights”) and by constructing a normative correlation matrix between the declaratory principles and the constitutional provisions. The results reveal significant convergences regarding human dignity, equality, and the repudiation of racism, as well as semantic gaps in the DPDHAIA and the absence of constitutional provisions on algorithmic governance. As a contribution, the study offers a discussion on the risk of AI reproducing racism, reinforcing prejudice, and fostering inequalities. It concludes that it is urgent to formulate public policies and legal frameworks for algorithmic regulation that incorporate a racial perspective, thereby ensuring that AI technologies are designed, developed, applied, and used with social justice and anti-racist responsibility.