U.S. Army pilots AI system to accelerate battlefield planning

Artificial intelligence is increasingly used in modern warfare, raising ethical questions about accountability when machine‑generated targeting errors occur. No international treaty yet governs autonomous weapons, and human oversight varies between “in‑the‑loop” and “on‑the‑loop” models.

The U.S. Army’s Combat Capabilities Development Command deployed an experimental generative‑AI platform called COA‑GPT during Project Convergence Capstone 6 at Fort Irwin, California (July 17‑29). Designed to compress mission‑planning timelines from days or hours to seconds, the system lets commanders generate and compare course‑of‑action options using a graphic interface that soldiers can manipulate on a shared map. Developers emphasize that the tool is intended to augment, not replace, soldiers, incorporating feedback from soldiers and industry partner Hyssos Tech. Future work will refine the platform through continued soldier testing and explore additional AI capabilities.