Attacco armato non intenzionale e legittima difesa putativa al tempo dell'Intelligenza Artificiale

The apparently unstoppable development of AI-powered early-warning systems and autonomous hack-back in the cyberspace raises the question as to whether existing rules and principles in the field of the law on the use of force (jus ad bellum) should be adapted to the reality of split-second decisions in critical fields, virtually lacking any meaningful human control. What if a States is mistakenly brought to perceive to be a victim of an armed attack and decides to adopt measures in self-defense pursuant to Article 51 of the United Nations Charter and the corresponding customary norms? May it invoke any right to act in “putative” self-defense? Is the norm prohibiting armed attacks construed as requiring a particular animus of aggression in addition to the long-standing requirement of the “gravity” of the use of force? The present article will appraise the constitutive elements of the category of “armed attack” based on the works of main UN bodies, State practice, international case-law and relevant scholarship, with a view to debating whether, although implicit so far, the requirement of hostile intention must be ascertained also in relation to the most grave uses of force characterizing as “armed attack”. The affirmative conclusion fits with a principled understanding of the rules and principles in this field and the actual way in which the systems under scrutiny will perform.