Algorithmic governance and power relations in hr beyond formal compliance

Abstract

This research critically analyzes whether the current regulatory framework — particularly the Brazilian General Data Protection Law (LGPD), the General Data Protection Regulation (GDPR), and recent AI regulatory initiatives — is sufficient to mitigate the structural ethical risks arising from algorithmic governance in Human Resource Management. The study demonstrates that, although AI is legitimized by narratives of efficiency and technical neutrality, algorithmic governance produces structural effects that impact privacy, equality, and the distribution of power in the workplace, shifting the debate from a purely technical dimension to a critical analysis of power structures and regulation. A systematic literature review was conducted in accordance with the PRISMA protocol, adopting a qualitative and analytical-critical approach and covering studies indexed in the Scopus and Web of Science databases between 2015 and 2025. The findings indicate that the main risks identified are not merely technical but structural: (i) the weakening of privacy due to the continuous datafication of workers and the limitations of consent in contexts of subordination; (ii) the reproduction of historical inequalities through algorithmic biases embedded in organizational data; and (iii) the intensification of power asymmetries resulting from informational centralization and decision-making opacity. The study found that the prevailing regulatory model remains centered on formal compliance, failing to adequately address the redistribution of power and the distributive impacts of automation. As a contribution, the article proposes the Structured Responsibility Algorithmic Governance Model (MGARE), based on workplace impact assessment, explainability, independent auditing, and collective participation, aiming to move beyond a documentary compliance paradigm toward governance oriented to substantive justice and expanded institutional accountability.