The U.S. military used the African Lion 2026 exercises in Morocco to field-test artificial intelligence across the battlefield, compressing decision cycles and showcasing robotic systems that could shift how wars are fought. Commanders employed Palantir’s Project Maven to ingest and analyze streams of sensor data, with operators querying information through Anthropic’s Claude—a sign the startup remains embedded in defense workflows despite recent political friction. Officers said decisions that once took hours were cut to minutes, while emphasizing a human still authorized lethal strikes.
The drills also highlighted emerging autonomy at the tactical edge. Overland AI demonstrated its ULTRA robotic vehicle, capable of navigating hazardous terrain and laying mines or explosives, with a remotely operated machine gun that could be automated in the future. Senior leaders acknowledged the ethical unease surrounding machines making lethal decisions even as they warned adversaries are pressing ahead and that autonomy may save soldiers’ lives by moving them off the front line. The exercise doubled as a live feedback loop for more than a dozen defense contractors competing to modernize U.S. forces.
Related article:
