ETD

• Autonomous Weapons Systems
• International Humanitarian Law
• Meaningful Human Control

The integration of AI into autonomous targeting systems exposes a profound accountability gap under international humanitarian law (IHL). While developers embed safeguards yet evade post-deployment liability, commanders risk over-reliance amid hallucinations and biases, and states face attribution challenges in classified operations. This research aims to dissect these diffused responsibilities across the AI lifecycle, proposing hybrid solutions: mandatory explainable AI disclosures for developers, meaningful human control (MHC) protocols for commanders, and national Article 36 of Additional Protocol I review boards for states. Drawing on recent cyber espionage abuses and wargame escalations (e.g., 95% nuclear tendencies), it highlights literature gaps in empirical cases and non-Western perspectives, urging norms aligned with the Convention on Certain Conventional Weapons (CCW) to bridge theory and battlefield realities.