Autonomous Combat Drone Race Accelerates as Boeing, Anduril and General Atomics Advance New Designs.

Boeing showcased its flight-tested MQ-28 Ghost Bat at the 2026 Farnborough International Airshow alongside competing designs from Anduril, General Atomics and Airbus, highlighting a growing race to field uncrewed combat aircraft. Boeing leads in publicly demonstrated flight and weapons testing, but Anduril and General Atomics hold the stronger near-term position through U.S. Air Force production contracts.

Reuters reported on July 22 that Lockheed Martin, Northrop Grumman, BAE Systems and autonomy-software companies are also pursuing a market that could demand hundreds of aircraft each year. The emerging requirement centers on affordable sensor, electronic-warfare and missile-carrying platforms that can extend the reach and survivability of crewed fighters under human control.

Related topic: U.S. DARPA Tests Sikorsky Nomad 100 Runway-Independent Drone for Long-Endurance Missions.

Boeing’s MQ-28 Ghost Bat at Farnborough highlights the accelerating race to field autonomous uncrewed combat aircraft able to carry missiles, sensors and electronic-warfare payloads alongside crewed fighters (Picture source: Army Recognition Group).

The MQ-28 remains the most mature aircraft in flight testing. Boeing lists a maximum take-off weight of 12,000 lb, a speed of up to Mach 0.9, a ceiling above 40,000 ft and range exceeding 2,000 nautical miles. Its interchangeable nose permits different sensor, electronic-warfare and communications payloads without redesigning the entire aircraft. Australia awarded Boeing an A$754 million three-year agreement in December 2025 covering seven additional aircraft, development and support; that figure should not be treated as a unit price because it includes engineering and test activity. Boeing reports more than 100 flights. In December 2025, an MQ-28 fired an AIM-120 AMRAAM and destroyed an airborne target during a mission involving an E-7A Wedgetail and F/A-18F Super Hornet.

The armament test was more significant than a conventional missile firing because it demonstrated a distributed engagement sequence. The F/A-18F contributed targeting information, the E-7A supported command and control, and the MQ-28 acted as the remote missile carrier. The AIM-120 is a 3.66-metre, approximately 151-kilogram beyond-visual-range missile with inertial mid-course guidance, an active radar seeker for terminal homing, and a blast-fragmentation warhead. This allows the launching aircraft to release the missile before its own radar has generated the complete engagement solution, provided targeting data and updates are available from another sensor. Moving the launch point forward improves missile kinematics by reducing the distance to the target, but the benefit depends on secure data links and accurate track correlation.

Inside the U.S. program, Anduril’s FQ-44 and General Atomics’ FQ-42 are ahead. On June 17, 2026, the Air Force awarded both companies engineering, manufacturing-development, and production contracts for Collaborative Combat Aircraft Increment 1, four months earlier than scheduled. The service plans to obtain more than 150 combat-capable aircraft by the end of 2030 and still identifies approximately 1,000 aircraft as its broader force objective. Its fiscal 2027 request includes $1.1 billion for CCA procurement. Anduril’s aircraft fired an AIM-120 at a digital target over the Mojave Desert on July 15 after earlier captive-carriage tests with inert weapons. The Air Force states that weapon release remains a human decision, although the aircraft can execute the firing sequence within operator-defined parameters. The externally carried missile used during testing also illustrates an unresolved design trade-off: external stores simplify integration but increase drag and radar signature.

General Atomics offers the longer institutional record in uncrewed aviation, but its FQ-42 program has also exposed the risks of compressed development. A test aircraft was lost shortly after take-off on April 6, 2026, when the autopilot used incorrect weight and centre-of-gravity calculations. Flight testing resumed on May 21 after software changes and a joint safety review. Before the accident, General Atomics had flown the aircraft using RTX Collins Aerospace’s Sidekick mission-autonomy software through the government-owned Autonomy Government Reference Architecture. This separation between flight controls, mission software and aircraft hardware is central to the Air Force’s acquisition strategy. Six companies, Anduril, General Atomics, Lockheed Martin, Northrop Grumman, RTX Collins Aerospace and Shield AI, hold mission-autonomy contracts, with a primary Increment 1 software supplier due to be selected by summer 2027.

The European field is less mature in flight testing. Airbus’ U760 Ravenstorm is a 13-metre-long aircraft with a 10-metre wingspan, designed for precision air-to-surface weapons, medium- and long-range air-to-air missiles, radar suppression and non-kinetic electronic attack. Airbus targets availability in the early 2030s. Its interim U740 Valkyrie combines two Kratos XQ-58-derived aircraft with Airbus’ MARS mission system, with flight tests planned in 2026 and a proposed German capability in 2029. BAE Systems unveiled Brontanax on July 22 as a British aircraft comparable in size to the Hawk trainer and intended for air combat, strike and electronic effects. BAE says more than 500 employees and 75 British suppliers are involved, but it has not released range, payload, engine or production dates. Boeing therefore leads Europe-facing competitors in demonstrated hardware, while Airbus and BAE retain advantages in sovereign European integration and domestic industrial participation.

Companies are pursuing these programs because the procurement structure can produce revenue well beyond aircraft assembly. The Air Force wants each CCA to cost no more than one-third of an F-35, whose flyaway price is approximately $80 million or higher, implying a target below roughly $27 million before some support, software and weapon costs. At an eventual inventory of 1,000 aircraft, acquisition alone would represent a market measured in tens of billions of dollars, followed by autonomy licences, sensor upgrades, communications equipment, munitions integration and sustainment. The competition is therefore also a contest over technical standards: a company whose autonomy software becomes widely adopted could earn revenue across several aircraft types even without winning the airframe contract.

Large-scale adoption would change air operations by separating the sensor, decision-maker and weapon carrier. A crewed fighter could remain outside the densest threat area while uncrewed aircraft search for emitters, jam radars, transmit targeting data or carry additional AIM-120 missiles. Commanders could place aircraft on several approach axes, compel an opponent to activate radars against decoys and preserve crewed fighters for missions requiring judgement or complex identification. These advantages are not automatic. Communications may be jammed, autonomous software must distinguish valid targets from ambiguous tracks, and larger fleets will require runways, maintainers, fuel, spare engines and weapons. The decisive measure will therefore not be whether these aircraft can fly autonomously, but whether air forces can generate them in useful numbers, connect them under electronic attack and replace losses at a cost consistent with sustained combat.

Explore More Defense News

• Land Defense News
• Naval Defense News
• Defense Aerospace News