EXECUTIVE SIGNAL
On 17 June 2026, the United States Air Force awarded production contracts to Anduril Industries and General Atomics for Collaborative Combat Aircraft Increment 1, the first time a major military has committed to serial manufacture of fighter-class autonomous combat aircraft. The service contracted over 150 FQ-44A Fury and FQ-42A Dark Merlin aircraft for delivery by the end of the decade, at a target unit cost below $30 million each, roughly one-third the flyaway price of an F-35A at recent production lots. Both designs were selected simultaneously rather than through a conventional down-select.
In a parallel award on the same date, Shield AI received a production contract for Hivemind mission autonomy software, designated as the primary autonomy layer under the Air Force's Autonomy Government Reference Architecture. A-GRA deliberately separates the software stack from the airframe, allowing the Air Force to upgrade autonomous behaviours independently of the physical aircraft on a different refresh cycle. Hivemind enables autonomous aircraft to reroute dynamically, execute collaborative tactics with crewed fighters, and respond to unexpected threats without continuous ground intervention.
The production decision was awarded four months ahead of the programme's internal schedule. The Air Force has stated a long-term objective of approximately 1,000 CCA aircraft, implying a total programme commitment at current target pricing of over $30 billion. At that scale, the CCA fleet would substantially expand the Air Force's tactical aviation inventory without a commensurate increase in aircrew numbers, which is the central logic of the crewed-uncrewed teaming concept the programme instantiates.
The selection of both airframe competitors rather than a single winner is itself a signal. The Air Force is preserving competitive pricing tension and supply-side redundancy across the production phase. For Anduril, the contract confirms it can deliver complex aerospace hardware on schedule. For General Atomics, a mature aerospace manufacturer, the win validates its repositioning as an autonomous systems company rather than a legacy ISR platform provider.
SIGNAL 01, THE DUAL-SELECTION LOGIC
The Air Force's decision to advance both the FQ-44A Fury and the FQ-42A Dark Merlin into production simultaneously departs from the down-select convention that has governed major aircraft competition. In a conventional competition, two competing designs converge on a single winner and the losing programme is terminated. The CCA decision instead treats both designs as viable and advances them together under a shared mission requirements specification and the A-GRA autonomy standard, creating a dual-source production programme for a new class of combat aircraft.
The dual selection introduces redundancy at the supplier level. A production disruption at either Anduril or General Atomics does not stop the programme. It also preserves pricing competition into the manufacturing phase: with two separate production contracts, the Air Force retains the ability to adjust procurement volumes between the two manufacturers at successive annual reviews, rather than being locked into the cost trajectory of a single source.
The Air Force FY27 budget request includes nearly $1 billion for initial CCA procurement, but the contracts themselves do not publicly disclose the split between the two manufacturers. The absence of disclosed contract values is consistent with the Air Force's stated preference for continuous competition as the primary cost-control mechanism throughout the programme, a preference the dual selection directly serves.
STRATEGIC IMPLICATION
If the dual-selection model is sustained through CCA Increment 2 and beyond, it would represent a structural break from the winner-takes-all platform competition that has defined US fighter procurement for four decades. It signals that the Air Force views CCA as closer to a commodity capability than a bespoke weapons system, with profound implications for how the programme manages cost and industrial risk across the full 1,000-aircraft objective.
STAY ON TOP OF THIS MARKET
Track this sector weekly with DI Pro.
Every Tuesday brief, every Thursday intelligence page, plus the Friday roundup of vendor moves, contract awards, and regulatory updates across the autonomous-systems sector.
SIGNAL 02, THE A-GRA SOFTWARE ARCHITECTURE
The Autonomy Government Reference Architecture establishes a formal boundary between the CCA's physical layer and its cognitive layer. Under A-GRA, the Air Force retains authority over autonomy software upgrades on a cycle that is independent of the airframe manufacturer's production schedule. Hivemind receives the current production contract as the initial software implementation, but the architecture itself is government-owned, meaning Shield AI is a contracted implementer rather than a permanent sole-source owner of the aircraft's decision-making capability.
The practical effect is that the Air Force can update mission autonomy behaviours, introduce new tactical profiles, or in principle replace the autonomy provider without modifying the aircraft. For the CCA programme, this eliminates a single point of software dependency on any one vendor. For the broader defence autonomy market, A-GRA establishes a reference certification standard that any future competitor to Hivemind would need to meet before being eligible for re-competition on the programme.
The architecture also has implications for allied interoperability. If partner air forces acquire A-GRA-compliant platforms, coalition CCA operations become technically feasible without requiring each nation to operate identical airframes or identical autonomy software. A common interface standard, rather than a common product, is the interoperability mechanism, a design principle consistent with how NATO has historically built coalition interoperability for communications and electronic warfare systems.
STRATEGIC IMPLICATION
The A-GRA framework transforms CCA from a traditional platform programme into a platform-plus-software programme with structurally separate competition tracks. If applied to future Air Force autonomous systems, as indicated is the service's intent, it would fragment the vertical integration advantage currently held by companies that bundle hardware and software in a single product. The competitive map for defence autonomy over the next five years will be shaped significantly by which companies achieve A-GRA certification first.
SIGNAL 03, THE FLEET ECONOMICS
The Air Force's unit cost target for CCA Increment 1 is below $30 million per aircraft, against an F-35A flyaway price of approximately $82 million at recent production lots. The target positions each CCA at roughly one-third the per-unit cost of the crewed aircraft it is designed to accompany. At the long-term objective of approximately 1,000 aircraft, the programme represents a potential commitment exceeding $30 billion at current target pricing, before lifecycle support and software upgrade costs.
The cost-per-effect comparison depends on mission assumptions. A CCA that absorbs threat exposure in place of a crewed aircraft, delivers additional ordnance, or executes suppression-of-enemy-air-defences tasks changes the risk-adjusted cost of a strike package materially. If autonomous wingmen routinely occupy the highest-attrition sortie positions, the effective loss-rate cost the Air Force bears per crewed aircraft falls, even before accounting for the removal of aircrew casualty risk from those positions.
The Increment 1 commitment of over 150 aircraft is a relatively small initial purchase against the 1,000-aircraft objective. The programme's continuous competition structure is designed to use early production data to drive cost reductions before the main procurement tranche. Whether unit costs track toward or below $30 million through the first production lot, and how that compares against the FY27 request of nearly $1 billion, will establish the programme's long-term credibility with Congress.
STRATEGIC IMPLICATION
The economic logic of CCA becomes most consequential in a high-attrition contested-airspace scenario. A fleet of 1,000 autonomous wingmen operating alongside the crewed force fundamentally changes air campaign planning: commanders gain tactical mass without commensurate personnel risk, and adversary planners must allocate defensive resources against a threat that cannot be deterred by the risk of aircrew casualties. The production order issued this week moves that calculation from analytical to operational.
DRONE INTELLIGENCE ASSESSMENT
The USAF CCA production awards of 17 June 2026 are the most consequential autonomous aircraft procurement decisions issued by any military to date. The simultaneous advance of two airframe designs under a shared software architecture and a government-retained autonomy standard is not an incremental programme milestone; it is the structural foundation of the next generation of American air power. The four-month acceleration against programme schedule indicates the Air Force has resolved the technology readiness questions it had expected to address during the current development phase, and the dual-selection model signals institutional confidence in the programme's commercial and operational assumptions.
The binding constraint over the next 12 to 18 months is operational rather than technical: the Air Force must develop tactics, training, and rules of engagement for crewed-uncrewed teaming before the first production FQ-44A and FQ-42A reach operational squadrons. The autonomous warfare architecture must also translate these procurement decisions into joint doctrine covering the Navy's and Army's parallel autonomous systems initiatives. If that doctrine development lags hardware delivery, the CCA fleet risks replicating at altitude the readiness gap that has constrained drone warfare at the tactical level since the Predator era.
CCA Increment 1 Programme Key Data
| Parameter | Detail |
|---|---|
| Award Date | 17 June 2026 |
| Airframe Producers | Anduril Industries (FQ-44A Fury), General Atomics Aeronautical Systems (FQ-42A Dark Merlin) |
| Autonomy Software | Shield AI Hivemind |
| Autonomy Framework | Air Force A-GRA (Autonomy Government Reference Architecture) |
| Increment 1 Quantity | Over 150 aircraft (combined, by end of decade) |
| Unit Cost Target | Below $30 million per aircraft |
| Long-term Fleet Objective | Approximately 1,000 aircraft |
| FY27 CCA Procurement Request | Nearly $1 billion |
| Schedule Status | 4 months ahead of internal programme plan |
CCA vs Selected US Air Combat Assets: Unit Cost Comparison
| Aircraft | Unit Cost (approx.) | Role | Crew |
|---|---|---|---|
| FQ-44A Fury (Anduril CCA) | Below $30M (target) | Autonomous combat wingman | Autonomous (human supervised) |
| FQ-42A Dark Merlin (General Atomics CCA) | Below $30M (target) | Autonomous combat wingman | Autonomous (human supervised) |
| F-35A (recent production lots, flyaway) | ~$82M | Multi-role fighter | 1 pilot |
| MQ-9B Reaper | ~$32M | ISR and strike | 2 operators (remote) |
SOURCES & REFERENCES
ABOUT THIS BRIEFING
- Prepared by
- Drone Intelligence editorial team
- Published
- Q2 2026
- Last verified
- 22 June 2026
- Sources
- 8 primary sources cross-checked
- Confidence
- High on verified facts. Assessment and forecast labelled inline.
- Corrections
- Email paul@droneintelligence.ai with the briefing URL and the source you believe contradicts the claim.
Prepared under the Drone Intelligence methodology. Editorial decisions follow our editorial policy. Independence and disclosure standards at ethics.
CITE AS
“The Production Order: USAF Commits to Fighter-Class Autonomous Wingmen and the Architecture of Crewed-Uncrewed Combat.” Drone Intelligence, Q2 2026. https://droneintelligence.ai/insights/usaf-cca-production-contracts
RELATED INTELLIGENCE
Drone Intelligence, Signal Dossier VOL. 02-U. Classified Distribution.
paul@droneintelligence.ai