USAF Marks First Autonomous Multi-Drone Combat Maneuvers With Shield AI and Anduril
Shield AI and Anduril Industries successfully demonstrated the U.S. Air Force’s first multi-drone autonomous air combat maneuvers at the Nevada Test and Training Range. Directed by Hivemind software, two FQ-44A drones executed tactical engagements under supervisory ground command.
The U.S. Air Force achieved a major operational milestone in its Collaborative Combat Aircraft (CCA) initiative on October 26, executing the program's first multi-drone autonomous air combat maneuvers. Conducted at the Nevada Test and Training Range, the flight featured two FQ-44A unmanned aircraft—built by Anduril Industries—performing beyond-visual-range tactical engagements coordinated entirely by Shield AI’s Hivemind mission autonomy software. Building upon an earlier October test where an F-22 pilot manually commanded a single asset, this flight represented a significant advancement: no human was in the sky. Instead, a ground-based operator provided only high-level supervisory directives as the paired drones autonomously executed complex tactical formations against a virtual threat.
This flight marks a pivotal transition from legacy tele-operated systems toward true multi-agent autonomous orchestration in uncrewed combat aviation. For defense leaders and operational commanders, the test offers initial empirical proof that algorithmic mission autonomy can successfully manage dynamic tactical engagements without requiring continuous, low-level human control. As the Air Force refines force structure models demanding crewed-to-uncrewed ratios from 1:2 up to 1:8, mission software must absorb the cognitive overload of flight mechanics, formation discipline, and threat evasion. Relieving cockpit crews of micro-management is mandatory if fifth- and sixth-generation platforms are to operate effectively within dense integrated air defense environments.
From a defense acquisition and industrial base perspective, the successful flight highlights the viability of software-centric, hardware-agnostic mission architectures. The seamless integration of Shield AI’s enterprise autonomy stack onto Anduril airframes demonstrates how non-traditional defense contractors can collaborate to rapidly field advanced combat capabilities outside traditional prime contractor cycles. Defense acquisition officials should expect future testing schedules to accelerate rapidly, moving beyond basic paired flights toward complex multi-team swarming operations, dynamic in-flight asset transfers between mission packages, and eventual live-weapons employment trials designed to validate end-to-end tactical lethality.
As the Department of Defense prioritizes low-cost combat mass to deter peer threats across contested theaters like the Indo-Pacific, multi-drone mission autonomy will serve as a core foundational pillar. Service leaders and defense prime partners must now focus effort on establishing resilient human-machine interface standards, ensuring robust datalink protection against electronic warfare, and institutionalizing rapid software deployment pipelines to operationalize these achievements. (Source: Air & Space Forces Magazine)
Executive takeaway
Subscribe to receive signals like this — plus the executive analysis behind them.
Get the Signals