Illustrative title card. ISC open-source assessment, 18 June 2026. A sourced image will accompany the published edition.

US Air Force Orders Anduril's FQ-44 Into Production in Landmark Collaborative Combat Aircraft Award

The Production Decision

On 17 June 2026 the United States Air Force (USAF) awarded production contracts for its first semi-autonomous Collaborative Combat Aircraft (CCA), moving Anduril Industries’ FQ-44 and General Atomics’ FQ-42 from prototype to manufacture. The two designs, known through flight testing as the YFQ-44A and YFQ-42A, drop the “Y” prototype prefix and become the FQ-44A and FQ-42A now that the service has committed to building them. Anduril’s aircraft, named Fury, is the subject of this assessment.

The award combines engineering and manufacturing development (EMD) with initial production. The service said the two companies will together build at least 150 combat-capable CCA by the end of the decade under the programme’s first increment, designated Increment 1. The contract covers the first three production lots. The service did not disclose the unit price or how orders will be split between the two firms. Separately, it selected three vendors, Anduril, Shield AI and Collins Aerospace (an RTX company), to compete on the mission autonomy software that will actually fly the aircraft.

We see CCA as representing the next evolution of airpower. When paired with our manned fighters, we can extend reach, increase survivability, and generate the mass that is necessary in combat in a highly contested environment. Col Timothy Helfrich, USAF Portfolio Acquisition Executive for Fighters and Advanced Aircraft, 17 June 2026

Anduril states the path from clean-sheet design to first flight took 556 days. The programme office describes the move from prototype award to production contract, just over two years, as the fastest path from prototype to production for a fighter aircraft in more than fifty years. The air-vehicle award came roughly four months ahead of schedule.

FQ-44: What Anduril Is Building

The FQ-44A Fury is a jet-powered, runway-launched uncrewed combat aircraft built to fly alongside crewed fighters under manned-unmanned teaming (MUM-T). Its intended teammates are the F-35A Lightning II, the F-22A Raptor and the planned sixth-generation F-47. The CCA concept hands the uncrewed aircraft roles such as air-to-air engagement, strike, reconnaissance, electronic warfare and jamming, taking cues from a pilot in an accompanying crewed jet rather than from a ground operator working a stick and throttle. On its first flight Anduril said the aircraft managed its own flight controls and throttle and executed a mission plan without an operator flying it remotely.

Weapons Integration and the Magazine Question

In February 2026 the Air Force released its first image of the Fury carrying an AIM-120 AMRAAM (Advanced Medium-Range Air-to-Air Missile), confirming the start of weapons-integration testing with a captive-carry round. Reporting indicates the aircraft carries stores on two external underwing hardpoints, with the longer-range AIM-260 JATM (Joint Advanced Tactical Missile) anticipated as a follow-on fit. Dispensing with an internal bay trades low observability with stores for manufacturing simplicity and lower cost, a deliberate choice for an aircraft meant to be built in quantity.

For the munitions community the significance lies less in the airframe than in the magazine it adds to the force. Every affordable uncrewed platform that can carry two or more air-to-air missiles puts extra launch rails into the fight at a fraction of the cost of a crewed fighter. In round numbers, Increment 1 alone (around 150 aircraft) would add roughly 300 external air-to-air missile stations to the force; a notional fleet approaching 1,000 tails, paired at about two CCA per crewed fighter, would mark a step-change in magazine depth and shooter count in contested airspace. With the programme office targeting a unit price below one-third of an F-35A, roughly under 30 million US dollars per tail, and an eventual fleet the service puts at around 1,000 aircraft, the CCA force becomes a substantial new claim on precision air-to-air missile stockpiles. Sustained operations would test AMRAAM and JATM production rates, and the energetics supply chains behind them, as much as aircraft output.

“Software Sold Separately”: A New Acquisition Model

The Air Force is deliberately separating the purchase of the aircraft from the autonomy software that flies it, an approach it calls “software sold separately”. Anduril, Shield AI and Collins Aerospace (RTX) each hold a six-month contract line item number (CLIN) to mature their mission autonomy toward the service’s initial operating capability (IOC) criteria. These three sit within a wider six-vendor baseline pool, which also includes General Atomics, Lockheed Martin and Northrop Grumman, retained on a six-year competitive contract built around the government-owned Autonomy Government Reference Architecture (A-GRA). The service will then narrow the field to one or two vendors for a further six-month option, with a single choice expected by summer 2027. The stated aim is to buy agile, updatable software that is not locked to a single airframe vendor, and to keep competitive pressure on both hardware and software through the programme’s life.

Analysis: Reach, Mass and Survivability

The decision matters on three counts. The first is pace. A prototype award in April 2024 leading to a production contract in June 2026 is, on the Air Force’s own account, the quickest prototype-to-production run for a fighter in over half a century, and a pointed contrast with the decade-long timelines of recent crewed programmes. The second is the dual-source outcome. By selecting both Anduril and General Atomics, the Air Force preserves two competing production lines, hedging against the single-line fragility that has repeatedly bitten Western munitions supply.

The third is the conditional nature of the “affordable mass” argument. Cheap, numerous aircraft only generate combat power if the weapons, datalinks and autonomy mature at the same rate as the airframe. A platform that cannot be armed, networked and trusted to act semi-autonomously at scale does not add mass; it adds hangar queens. The separation of software from hardware is the Air Force’s bet that it can keep all three strands moving together, but it also creates an integration responsibility that someone has to own for the life of the fleet.

Programme Risks and Data Gaps

Several parameters remain closed or unverified. Unit price is classified; the Air Force confirms only a target below one-third of an F-35A. The division of the first three lots between Anduril and General Atomics is undisclosed. The performance figures above (engine, thrust, ceiling, radius) rest on open-source reporting rather than official release and should be treated as indicative. The autonomy competition is unresolved, so IOC timing depends on a down-select not due to conclude until summer 2027. Programme risk was illustrated by the April 2026 crash of a General Atomics YFQ-42A shortly after take-off in California, which paused that aircraft’s flight testing for about six weeks before resuming with a software fix; the service said the incident did not affect its decision to back both designs. Looking further out, the harder problems are less about the airframe than the system around it: datalink resilience in contested electromagnetic environments, certification of the rules under which a semi-autonomous aircraft may employ weapons, and long-term ownership of the hardware-and-software integration seam the separated acquisition model creates. Increment 2 planning is expected to explore more capable CCA variants with greater range, payload or sensing. None of these gaps alters the central fact: the FQ-44 is now a production aircraft.

Corrections & updates welcome. If you hold open-source data that refines or corrects any parameter in this article, please contact [email protected] citing the specific claim and your source. Verified corrections will be incorporated and credited in the revision history. AI-assisted technical assessment based on open-source material. Not a formal intelligence product.