Gobernanza algorítmica y democracia: desafíos éticos de la inteligencia artificial en el estado

Abstract

This study examines the social, ethical, and legal implications of using artificial intelligence (AI) tools in the administration of social services, particularly in the identification and management of vulnerable populations. Through a mixed-methods approach, the research explored both the perceptions of key social actors and the concrete impact of algorithms used in automated systems for evaluating beneficiaries. The findings show that although AI can enhance operational efficiency and enable faster resource allocation, it may also reinforce structural biases and lead to algorithmic discrimination, especially affecting historically marginalized groups. The study identifies risks such as system opacity, limited institutional accountability, and insufficient citizen participation in the design and implementation processes of these technologies. The discussion highlights the urgent need for regulatory frameworks that ensure ethical AI aligned with human rights and principles of social justice. The development of algorithmic impact assessment tools, along with strategies for transparency, public engagement, and democratic oversight, are proposed as key pillars for fair AI governance in the field of social protection. The conclusion emphasizes that without a human-centered, inclusive, and regulated approach, the deployment of these technologies may exacerbate existing inequalities and seriously compromise equitable access to basic rights and services. A collective and interdisciplinary effort is essential to achieve the ethical integration of AI into social policy.