Following the end of the joint NGF, Europe continues its Combat Cloud. Architecture, drones, industrial base, and the acceptable cost of attrition are decisive.
Executive Summary
The termination of the joint NGF fighter in June 2026 did not cause the complete demise of FCAS. France and Germany confirmed in July their commitment to pursue the Combat Cloud, while work on collaborative drones remains technically usable. The core concept is to interconnect fighters, drones, sensors, satellites, and effectors in a distributed architecture capable of sharing and processing data in near real-time. However, a central question arises: how much can a drone cost if we accept its loss? FCAS mixed several categories, ranging from small expendable Remote Carriers to reusable stealth drones. Yet these platforms operate under fundamentally different economic models. A drone costing a few million euros can reasonably be exposed to high risks. A stealth platform costing tens of millions effectively becomes an uncrewed combat aircraft that must be protected. Europe must now define this economic boundary before finalizing its military architecture.
FCAS Survives the Cancellation of Its Joint Fighter
June 8, 2026, fundamentally altered the nature of the Future Combat Air System. France and Germany abandoned joint development of the New Generation Fighter following years of conflict between Dassault Aviation and Airbus over industrial leadership and governance.
However, the remainder of the program was not abandoned.
During the Franco-German Council of Ministers on July 17, Friedrich Merz confirmed that the cloud solution originating from FCAS would proceed “within a different framework.” Emmanuel Macron indicated that Combat Cloud projects would move forward according to the planned schedule.
The initial program had already absorbed €3.2 billion for Phase 1B. The contract signed in 2022 could have reached nearly €8 billion including the optional Phase 2. This structure must now be reorganized given that the joint NGF demonstrator—a central pillar of the effort—no longer exists.
This rupture extends the challenges analyzed surrounding European air sovereignty and fighter development. Yet paradoxically, it shines a clearer spotlight on the most modern segment of the program: the network.
The Combat Cloud Is Not a Simple Communications Network
Referring to a “cloud” can be misleading. It is not about routing all military data to a massive central server.
The targeted architecture is distributed.
A Rafale, an Eurofighter, a drone, an airborne early warning aircraft, a satellite, or a frigate each become nodes within the same network. Every platform contributes information while processing data locally.
The operation relies on multiple layers. Sensors detect a threat. Data is fused. Artificial intelligence assists in prioritization. Relevant information is distributed to users who require it. The system can then assign a task to the most favorably positioned effector.
For example, a Rafale might partially detect a surface-to-air defense system without precisely identifying it. A Remote Carrier positioned closer completes the targeting location. A third platform conducts electronic jamming. A missile launched by yet another aircraft could ultimately engage the target.
The sensor and the shooter no longer need to be the same platform.
Airbus historically led the Combat Cloud pillar, alongside Thales for France and Indra for Spain. Its MARS architecture illustrates the current technical direction: platform-independent software, open interfaces, distributed processing, edge computing, and multi-domain data exchange spanning air, land, sea, and space.
FCAS Drones Do Not Fall Into a Single Category
This is where the attrition debate becomes critical.
The term Remote Carrier encompasses vastly different platforms.
MBDA is working on an Expendable Remote Carrier measuring approximately 4 meters in length and weighing 400 kg, flying at subsonic speeds for roughly one hour. It could carry a jammer, a sensor, a communications relay, or act as a decoy.
Its primary design premise accepts its destruction.
At the opposite end of the spectrum is a new generation of full-fledged uncrewed combat aircraft. Airbus is developing the U760 Ravenstorm, measuring approximately 13 meters long with a 10-meter wingspan. It is intended to execute strike missions, engage in air-to-air combat, or conduct electronic warfare.
Positioned between the two, Airbus plans to field the U740 Valkyrie with the Luftwaffe as early as 2029, based on the American Kratos XQ-58 and powered by the European MARS mission system.
The A400M could even serve as a mothership platform: Airbus estimates it could air-launch up to 50 small Remote Carriers or 12 heavy platforms.
This fundamentally alters available mass in combat scenarios.

Acceptable Cost Defines the True Meaning of “Attritable”
An attritable system is not necessarily disposable. It is engineered to return, but its occasional loss must remain economically tolerable if it protects a crewed aircraft or ensures mission accomplishment.
The distinction is critical.
A €5 million drone dispatched against an air defense system to protect a fighter and its pilot represents a rational exchange. A €30 million or €40 million drone featuring an AESA radar, an advanced engine, and a complex electronic warfare suite can no longer be sacrificed lightly.
Europe currently publishes no official cost ceiling for FCAS Remote Carriers.
U.S. programs nevertheless provide useful benchmark points. Kratos recently cited a unit cost of approximately $5.5 million for a low-rate production version of the XQ-58 Valkyrie, while projecting roughly $4 million for an order of one hundred units in a different configuration.
At the higher end, the U.S. Navy established a maximum target of $30 million per carrier-suited Collaborative Combat Aircraft in 2026.
This cost span demonstrates that “combat drone” is no longer a precise economic category.
The Real Cost Includes the Payload
Setting a price tag for the airframe alone is insufficient.
A Remote Carrier may feature a relatively low airframe cost while carrying millions of euros in onboard electronics. Radars, infrared sensors, jammers, secure datalinks, and munitions rapidly escalate the actual value exposed to risk.
Consequently, the full attrition cost must be calculated: airframe, sensors, weapons, software, support, and crucially, the industrial capacity to replace losses rapidly.
A €5 million drone manufactured at a rate of ten units per year is less “attritable” than an €8 million platform produced at a hundred units per year.
Attrition is as much an industrial factor as a financial one.
This is perhaps the primary lesson from Ukraine. Mass does not depend solely on a low unit price; it demands an industrial supply chain capable of replenishing inventory.
France Concurrently Develops a Far More Valuable Drone
French strategy incorporates an important nuance.
The future loyal wingman associated with the Rafale F5 operational standard will leverage operational lessons from the nEUROn technology demonstrator. It is intended to be stealthy, feature internal weapons bays, carry advanced sensors, incorporate embedded AI, and support aerial refueling.
This is clearly not a low-cost expendable platform.
Legislation updating France’s Military Programming Law in August 2026 plans for initial escort drone experimentation around 2028, leading to an operational stealth combat drone demonstrator by 2035.
France will therefore likely deploy a multi-tiered drone inventory: relatively simple expendable effectors, reusable Remote Carriers, and a significantly more costly stealth UCAV operating alongside the Rafale F5.
The Combat Cloud Can Link Aircraft Europe No Longer Builds Together
This may be the most compelling outcome following the end of the NGF.
France could develop its own successor to the Rafale. Germany can pursue an Airbus-led solution, join another multinational effort, or purchase additional American aircraft. Meanwhile, the GCAP initiative offers an alternative fighter project.
Europe could therefore field several distinct fighter platforms in the future.
What must remain common is no longer necessarily the airframe itself. It is the digital language enabling disparate aircraft to fight together.
A genuinely open European Combat Cloud could connect the Rafale F5, Eurofighter, future airframes, drones, and air defense systems without mandating a single standardized airframe.
However, this requires resolving questions far more sensitive than aerodynamics: Who holds the cryptographic keys? Who owns the operational data? Which nation authorizes software updates? Which algorithms assign target engagement to a drone? What occurs when satellite communications are jammed?
The next decisive test for FCAS may not involve manufacturing an aircraft. It will be proving that multiple nations can share a sovereign digital combat architecture while retaining national command over their weapons.
And for uncrewed platforms, the equation will be even more stark: Europe must decide in advance how much it is prepared to lose. If every Remote Carrier becomes too sophisticated, too scarce, and too costly to sacrifice, Europe will simply have created another expensive fighter—without a pilot inside.
