From Algorithmic Risk to Irreversible Harm

High-Risk Artificial Intelligence, Guarantor Position and Armed Conflicts

Abstract

This article examines the legal challenges arising from the use of high-risk artificial intelligence systems in armed conflict scenarios through the lens of legal risk theory, the guarantor position doctrine, International Humanitarian Law and International Human Rights Law. Using a qualitative, documentary and doctrinal methodology, the study analyses the concepts of risk, threat, danger and harm, as well as their implications for military decision-making supported by algorithmic systems. The article argues that the identification of foreseeable algorithmic risks generates legal duties of prevention aimed at avoiding their transformation into harms incompatible with the principles of humanity, distinction, proportionality and precaution. It further explores the relevance of human dignity, meaningful human control and the guarantor position as legal categories applicable to emerging technologies. The study concludes that existing legal frameworks provide sufficient tools to address the challenges posed by artificial intelligence in armed conflicts and that States have a responsibility to adopt regulatory measures aimed at preventing risks and protecting fundamental rights.