The deployment of artificial intelligence (AI) systems, particularly those employing machine learning (ML), introduces significant complexities in allocating responsibility when these systems fail, especially in high-stakes contexts such as military operations. This chapter examines how State obligations under international law, particularly due diligence obligations, intersect with the challenges posed by AI failures. It explores the blurred lines between human errors and technical malfunctions, highlighting how the inherent unpredictability of AI systems challenges traditional legal concepts such as negligence, risk, foreseeability, and reasonableness. The analysis addresses the lifecycle of AI systems, specifying the content of due diligence obligations across the stages of design, development, adoption, and deployment. It highlights the need for rigorous technical standards, robust testing, and comprehensive risk assessments to ensure compliance with international law. By rethinking the application of due diligence obligations in light of AI’s unique characteristics, the chapter provides a framework for States to fulfil their legal obligations while addressing the novel risks posed by AI technologies.