Legal, Ethical, and Technical Dilemmas in Military Artificial Intelligence

Edited by Bérénice Boutin, Taylor Kate Woodcock and Sadjad Soltanzadeh, this volume gathers the output of the DILEMA project at the Asser Institute in The Hague, a Dutch Research Council-funded effort to pull international lawyers, philosophers of technology and computer scientists onto the same page about military artificial intelligence. Published open access by T.M.C. Asser Press in April 2026, it runs to thirteen chapters across 336 pages.

Its central argument is that military AI cannot be treated as a single object. The editors split the field into three strands — autonomous weapon systems that select and engage targets, decision-support tools that shape what commanders see, and AI that augments the human soldier directly — and insist that each raises its own questions of agency, responsibility and control. Governance, the book contends, has to catch up on all three fronts at once, through legal compliance regimes, ethically informed technical design, accountability frameworks and clear normative limits on where AI should not be allowed to go at all.

The chapters move from theory to code. Klaudia Klonowska and Taylor Kate Woodcock take apart the rhetorical move of comparing machines to humans in legal debates. Marta Stroppa applies targeting law to autonomous cyber capabilities. Elisabeth Hoffberger-Pippan asks what an Article 36 weapons review looks like when the system under review is a large language model like ChatGPT. Thomas Bächle examines the AI-augmented super soldier through the lens of the human-machine interface, while Tine Molendijk, Sofie van der Maarel and Lonneke Peperkamp trace the moral-psychological pressure on combatants as autonomy in the weapon grows. On the technical side, Iris Cohen, Gregor Pavlin and Rianne Gouman work through context-driven requirements extraction, Jonathan Kwik proposes iterative assessment of military AI systems, and Tomasz Zurek and Tom van Engers tackle the hard problem of computationally representing rules of international humanitarian law inside autonomous devices. Governance chapters by Sofia Romansky, Tim Sweijs, Marcel Schliebs, Vanessa Vohs and Boutin herself turn to the quintessential role of responsibility, national legislation and industry self-regulation as a way out of the stalemate in autonomous weapons regulation, and the question of who answers when an AI system fails somewhere between negligence and malfunction.

Among the growing shelf on autonomous weapons, the volume is distinguished by its deliberate interdisciplinarity, setting the lawyer’s question, the ethicist’s question and the engineer’s question side by side. It is written for scholars, policymakers, military professionals and technologists, and because it is open access, every chapter is free to read.