Episode 37: Algorithms Deciding Who Lives and Dies

In an era of "decision advantage," military AI promises to process vast quantities of data to identify targets at a pace no human can match. But as we accelerate the "targeting cycle," are we accidentally engineering human reflection out of the loop? This episode explores the tragic real-world consequences of automated conflict—from the strike on a girls' school in Iran to the death of Abdul-Rahman al-Rawi, the first publicly acknowledged victim of AI-assisted targeting. We analyze why "meaningful human control" is failing and how the design of these systems "nudges" commanders toward certain decisions before they even see the data. Key Takeaways

• The Illusion of Speed: Why "faster" doesn't always mean "better." We discuss how AI-generated recommendations produced at a pace humans struggle to interrogate can turn a commander's judgment into a mere procedural check-box.
• Case Study: Abdul-Rahman al-Rawi: A deep look at the 2024 strike in Iraq that moved the debate from the theoretical to the identifiable.
• The Invisible Decision Environment: How AI shapes what information is prioritized and which options appear most "credible," effectively narrowing the range of independent judgment.
• The Lifecycle of Accountability: Why governance must start years before a strike—focusing on training data, algorithmic assumptions, and institutional incentives that currently reward speed over accuracy.
• Practical Governance: Moving beyond abstract UN debates to ensure operators have the training, time, and authority to critically evaluate—rather than just approve—AI outputs