This whitepaper articulates the dual relationship between agentic artificial intelligence and the twenty-first-century energy transition: on the one hand, AI is one of the most powerful tools for optimizing power grids, integrating renewables, modeling climate with unprecedented precision and accelerating decarbonization. On the other hand, the compute infrastructure of contemporary models generates a growing energy and carbon footprint. Current deployments in smart grids, industrial optimization and regional climate modeling are analyzed, and the net balance is quantified considering both vectors. An Ibero-American roadmap is proposed so that the region — with its relatively clean energy matrix, water abundance and minerals for transition — leads the model of energy-responsible AI.
Chris Meniw (2026) studied this question.