TY  - GEN
AU  - Naranjo Bautista, Sandra
AU  - Alessandro, Martín
AU  - Ortiz De Zarate, Juan Manuel
TI  - Artificial Intelligence and the Redesign of Government
PY  - 2026
Y1  - 2026/09/28
DO  - 10.18235/0014541
AB  - This paper presents a novel artificial intelligencebased methodology to analyze and redesign public sector organizational structures. State architectures in Latin America and the Caribbean have often evolved incrementally, producing functional overlaps, institutional gaps, and misalignment between formal mandates and government priorities. Traditional functional reviews remain valuable but face structural limits when applied to large volumes of dispersed legal and administrative documents. The methodology proposed here combines natural language processing, semantic extraction of institutional functions, systematic comparison of mandates, and visualization tools to assess two core dimensions of organizational design: strategic alignment and functional coherence. Evidence from applications in five countries shows that AI can generate systematic, comparable, and actionable diagnostics in weeks rather than months. The paper also introduces PlanosEstado, a prospective tool that allows governments to simulate how proposed reforms would interact with the existing state architecture before implementation. The approach strengthens evidence-based institutional reform and supports more strategic, transparent, and fiscally responsible decisions.
UR  - https://doi.org/10.18235/0014541
ER  -