Can AI autonomously coordinate swarm attacks using purely insect-scale drones in urban environments ?

Autonomous swarms of insect-scale drones operating in cities would require fleets of tiny, coordinated robots capable of complex tactics without human control. What progress has actually been made toward such systems, and where do current capabilities fall short of fully autonomous attacks?

Background

Recent miniaturization and advances in swarm intelligence have produced insect-scale drones capable of coordinated flight, but integration into real-world urban combat remains speculative. Current autonomous coordination platforms are limited to research prototypes demonstrating basic formation flight or obstacle avoidance rather than offensive maneuvers. Harvard Microrobotics Lab and the University of Washington’s Autonomous Flight System Lab have demonstrated millimeter-scale robots executing collective mapping and collision-free navigation in cluttered indoor environments, yet these systems depend on external motion capture or GPS-denied indoor positioning systems for localization and tracking. None of these implementations has demonstrated robust, real-time adversarial planning or fully autonomous coordination sufficient for hostile deployment in urban environments. As of May 10, 2026, no peer-reviewed evidence supports the existence of fully autonomous, purely insect-scale drone swarms executing coordinated offensive maneuvers in real urban settings.