L’attaque de l’Iran par les États-Unis : une guerre pour l’image ?

À partir d'Arendt, des problem-solvers à l’intelligence artificielle

Abstract

Since the U.S.-Israeli strikes on Iran began on 28 February, 2026, the Trump administration has repeatedly proclaimed victory, although none of its publicly stated war aims appears to have been met. This article asks how such a gap between discourse and reality can be understood. Drawing on Hannah Arendt’s analysis of the Pentagon Papers in "Lying in Politics", it proposes, as an interpretive hypothesis, that the war is driven by the production of an image of might – a politics in which factual truth becomes superfluous and decisions are detached from reality. It extends Arendt’s figure of the "problem-solvers" to artificial intelligence: by proposing scenarios to decision-makers, AI performs an analogous but radicalized function of derealization along four dimensions: probabilistic decision-making, structural opacity, the reduction of persons to numbers, and the operator’s alienation from their own decisions. It finally opens up to the idea that, although the structure is analogous, the underlying content of the image of might may strongly differ from that of the Vietnam-era administrations.