With the FQ-42 Vengeance and FQ-44 Fury, the USAF is preparing 1,000 CCA drones to increase the mass, range and survivability of its fighter aircraft.
In brief
The U.S. Air Force has now moved Collaborative Combat Aircraft from the prototype stage to the operational force. General Atomics and Anduril have been selected to produce the FQ-42 Vengeance and FQ-44 Fury. Increment 1 is expected to involve at least 150 aircraft by the end of the decade. The announced objective is now much broader: approximately 1,000 CCAs, with at least 500 in service by 2032. These semi-autonomous combat drones will not directly replace the F-35, F-22 or future F-47. They are intended to multiply their sensors, missiles and avenues of attack while keeping pilots farther away from the most dangerous areas. In this way, the USAF aims to regain a level of mass that has become financially impossible with manned fighters alone. But the change will be profound: new units, new maintenance requirements, autonomy software independent of the airframes, and new human-machine combat tactics.
The CCA program is now moving from prototype to production
The turning point came on June 17, 2026. The U.S. Air Force decided to retain both finalists in its Collaborative Combat Aircraft program. General Atomics will produce the FQ-42 Vengeance. Anduril Industries will build the FQ-44 Fury.
Increment 1 calls for at least 150 CCAs combining the two models before the end of the decade. The scale will then change again. In September 2026, Secretary of the Air Force Troy Meink set the objective of having at least 500 aircraft in service by 2032, as part of a target fleet of approximately 1,000 CCAs.
This is therefore no longer an experimental program.
The FY2027 budget proposal requests $996.5 million to begin procurement, $150 million in advance funding for 2028, and approximately $1.37 billion to continue development. The CCA program thus represents nearly $2.37 billion in 2027.
This expansion extends the strategy of the Loyal Wingman in American air warfare, but with a much greater industrial ambition.
The American strategy is to recreate mass without multiplying the number of pilots
The USAF’s problem is mathematical. A modern F-35 costs tens of millions of dollars, requires a pilot who takes a long time to train, and depends on extensive maintenance infrastructure. The future F-47 will be even more valuable.
It is therefore financially impossible to meet numerical requirements with a fleet composed solely of high-end manned fighters.
CCAs introduce a different equation. The initial objective was to keep their cost at around one-third that of an F-35, or approximately $30 million or less. Program officials stated in spring 2026 that they had fallen below this target, although no definitive contractual price per aircraft has been published.
The CCA thus becomes a form of affordable combat mass.
An F-35, F-22 or F-47 could be accompanied by several drones. The USAF’s historical assumption was two CCAs per manned fighter. Experiments now indicate that some crews could supervise three to five aircraft, or even more depending on the mission.
However, a distinction must be made. The pilot does not control each drone remotely like an MQ-9 operator. The pilot provides objectives and constraints. The CCA then manages part of its navigation, positioning, sensors and tactical behavior.
CCAs position weapons and sensors ahead of the fighters
The operational advantage lies first and foremost in the geometry of combat.
A CCA can fly several dozen or several hundred kilometers ahead of a manned fighter. It can observe an area using its own sensors, transmit information, search for a threat or carry additional missiles.
The pilot’s position consequently becomes less revealing.
This architecture also makes it possible to multiply avenues of attack. Instead of detecting four F-35s approaching from one direction, an adversary might have to deal with several fighters and a dozen CCAs dispersed throughout the airspace. Some may carry weapons. Others may conduct reconnaissance, jamming or serve as forward sensors.
The desired result is a saturation problem. An air-defense system has a finite number of radars, missiles and processing capabilities. CCAs allow the USAF to present more simultaneous threats without exposing as many pilots.
This logic directly complements the development of the F-47 and its networked combat architecture.
Operational deployment will change fighter squadrons
The real challenge begins now.
A fleet of 500 CCAs cannot simply be added to existing units. The Air Force must organize their maintenance, stocks of engines and spare parts, weapons, communications, ground crews and integration with fighter squadrons.
The USAF is already testing this organization at Creech Air Force Base, Nevada. In July 2026, FQ-42s and FQ-44s took part in Agile Combat Employment operations. Teams worked on weapons loading, sortie generation and operating the drones from an operational base rather than a test center.
The next step is scheduled for December 2026. The CCAs are to participate in Emerald Flag at Eglin Air Force Base, Florida. This will be their first integration into a large-scale air-combat exercise with fourth- and fifth-generation aircraft.
This is decisive. A high-performance drone that requires thirty technicians, specialized equipment and several hours of preparation between flights does not produce the desired mass.
Logistical simplicity therefore becomes almost as important as stealth or autonomy.
An open architecture makes it possible to change the brain without changing the aircraft
The most important innovation may be invisible.
The U.S. Air Force has deliberately separated airframe development from autonomy-software development. General Atomics and Anduril are building the aircraft. But Anduril, Shield AI and Collins Aerospace are competing separately to provide the software needed to execute missions.
This strategy is based on the Autonomy Government Reference Architecture, or A-GRA. The U.S. government controls the main interfaces. The software supplied by one provider must therefore be able to operate on another provider’s aircraft.
The concept has already been demonstrated. The FQ-42 has flown with Collins Aerospace’s Sidekick software. The FQ-44 has tested different autonomy systems and has even changed mission software during a single test flight.
This so-called “software sold separately” approach is intended to prevent one manufacturer from controlling the aircraft, its software and all of its future upgrades at the same time.
Above all, it enables much faster modernization. In theory, it becomes possible to improve the tactical behavior of an entire fleet without physically modifying the aircraft.
The technical operation of this architecture is explained in the CCA program’s transition to industrial production.
The human factor remains central to weapons employment
The term “autonomous drone” may suggest that the aircraft will decide on their own when to strike. That is not the framework adopted by the Department of Defense.
DoD Directive 3000.09 requires autonomous or semi-autonomous systems to allow commanders and operators to exercise an appropriate level of human judgment over the use of force.
In practice, the real shift concerns less the authorization to fire than everything that comes before it.
The CCA can manage its flight, trajectory, formation, certain data-processing tasks and positioning in relation to threats. The human pilot must therefore move from the role of remote operator to that of mission commander for a group of aircraft.
This represents a major transformation in cognitive workload. A pilot will have to understand what several aircraft are seeing and doing while continuing to fly their own fighter and engage in combat.
The USAF must therefore develop the aircraft, human-machine interfaces and doctrines needed for pilots to trust them simultaneously.
CCAs give the USAF an advantage particularly suited to the Pacific
The Chinese scenario gives the program its full significance.
In the Pacific, distances are immense. American bases are relatively few and potentially exposed to missiles. Fighters must cover vast areas while facing air-defense systems, fighter aircraft and long-range missiles.
Sending only F-35s and future F-47s into this environment concentrates value in a few highly expensive platforms.
CCAs, by contrast, make it possible to distribute combat power.
A commander will be able to preserve manned aircraft while sending sensors and weapons forward. The commander will also be able to accept greater risks with certain CCAs. The USAF now describes them as “optionally attritable” aircraft: they are not designed to be disposable drones, but their loss may be accepted when the military objective justifies it.
This distinction is fundamental. A CCA costing several tens of millions of dollars is not a consumable munition. But losing a drone is still militarily very different from simultaneously losing a next-generation fighter and its pilot.
Success will depend less on artificial intelligence than on scaling up
The technology has already passed several important milestones. The FQ-42 has been flying since August 2025. The FQ-44 achieved semi-autonomous flight after only 556 days of development. Software from different suppliers is being tested on multiple airframes.
The difficulties have not disappeared, however. In April 2026, a test FQ-42 was destroyed after its automatic flight-control system made a mass-and-balance calculation error. Flights resumed six weeks later after the software was corrected.
This incident perfectly summarizes the challenge. Building a drone capable of flying on its own is one thing. Operating 500 aircraft in a combat force is another.
The USAF’s potential advantage therefore lies not only in the FQ-42 Vengeance or FQ-44 Fury. It lies in the combination of hundreds of platforms, continuously improved software, manned fighters and an industry capable of producing aircraft rapidly.
If this architecture works, CCAs will change the very definition of air power. The fighter will no longer necessarily be the basic unit of combat. It will become the center of a small, distributed formation of sensors, weapons and autonomous aircraft. Quantity will once again become a military quality. flyajetfighter