Un registro de algoritmos, el primer paso para una IA transparente en España | lamarea.com

This article warns that the growing use of algorithms in public administration may violate fundamental rights by making unfair decisions without providing citizens with understandable explanations. The opacity of these systems, which range from the denial of social benefits to border control, creates a democratic deficit in which citizens lack the means to question or understand the automated criteria that affect their access to essential services. Genuine transparency is urgently needed to ensure accountability and prevent systemic discrimination. Although European regulation establishes registries for high-risk AI systems, it presents critical vulnerabilities that limit its effectiveness. By excluding algorithms deemed less “high-risk” that directly impact basic rights, concealing security systems in private areas, and applying the rules only to new developments, current legislation fosters superficial transparency. This insufficiency prevents real external oversight, leaving administrators and society without the tools to detect biases or errors in processes that are already operational. Regarding open data, the text emphasizes that merely publishing data is insufficient; accessible, complete, and up-to-date information is required for all automated systems, including those that have been decommissioned. The key point lies in the need to transform passive registries into dynamic tools for public scrutiny, mandating obligations and clear standards that enable civil society, the press, and researchers to critically audit algorithmic decision-making. Only in this way can we ensure that technology serves justice rather than undermining it.

Source: lamarea.com
Published on 2025-03-20

Un registro de algoritmos, el primer paso para una IA transparente en España | lamarea.com

Related news