The Pentagon aims to counter China’s numerical superiority with thousands of autonomous drones. Replicator promises speed, but poses real risks.
In summary
The Replicator program is not specific to the USAF. It is a Pentagon initiative, launched in 2023, to accelerate the production and deployment of disposable autonomous systems in several domains: air, sea, subsurface, and counter-drone. Its objective is clear: to counter China’s military might not through a symmetrical arms race involving aircraft carriers, destroyers, and stealth fighters, but through thousands of cheaper systems that are more numerous and faster to produce. The idea is appealing. It aligns with lessons from Ukraine, the Red Sea, and the Middle East: inexpensive drones can overwhelm very costly defenses. But Replicator does not spell the end of high-value platforms. It forces them to evolve. F-35s, destroyers, submarines, and satellites remain indispensable. The real breakthrough lies in the combination of rare platforms and distributed swarms.
The Replicator program is not a weapon, but a method
The name is misleading. Replicator is not a single drone. It is not a missile. Nor is it a traditional USAF program. It is an accelerated acquisition and deployment method led by the Department of Defense, with a key role assigned to the Defense Innovation Unit.
The initiative was announced in August 2023 by Kathleen Hicks, then Deputy Secretary of Defense. The message was clear: the United States must deploy several thousand autonomous systems across multiple domains within eighteen to twenty-four months. The key word is mass. The Pentagon believes that China holds a quantitative advantage in several categories: ships, missiles, personnel, industrial capacity, and logistical depth. Washington cannot respond solely with larger, more expensive platforms that take longer to produce.
Replicator is therefore based on a simple idea: produce faster, in greater numbers, with systems inexpensive enough to be lost in combat. In U.S. military parlance, these are called “attritable” systems. This does not mean disposable in the crude sense. It means that the loss of a system is acceptable within the operational calculus.
This logic profoundly changes the way we think about military power. An Aegis destroyer, an F-35, or a B-21 bomber are scarce assets. They are expensive, take a long time to produce, and are politically sensitive to lose. An autonomous reconnaissance drone, a loitering munition, or a surface drone can be deployed with a higher risk tolerance. This is the core of the Replicator program.
The Pentagon’s strategy targets China above all
Replicator was not conceived in a vacuum. Its implicit adversary is China. More specifically, the scenario haunting the Pentagon is the Indo-Pacific: Taiwan, the first island chain, immense distances, vulnerable U.S. bases, limited missile stocks, and the rise of the Chinese navy.
In this theater, scale is a physical problem. Millions of square kilometers must be monitored.
Sensors must be dispersed. Long supply lines must be maintained. Enemy targeting must be complicated. China must also be forced to expend costly missiles against numerous, less valuable targets.
This is where Replicator comes into play. Surface drones can patrol straits. Air drones can provide tactical intelligence. Loitering munitions can threaten vehicles, radars, or light vessels. Unmanned underwater systems can monitor or disrupt naval activity. Counter-drones can protect bases.
The Pentagon isn’t just looking to buy drones. It’s looking to create a short loop between military need, industry, software, production, and deployment. This is an implicit critique of the traditional American model, where a major program can take fifteen years to reach full maturity.
To be fair, this traditional model isn’t going away. It produces exceptional capabilities. But it’s too slow in the face of drone warfare, which can evolve in a matter of weeks.
The initial capabilities demonstrate a very pragmatic approach
Replicator’s first public announcements confirmed this pragmatic approach. AeroVironment’s Switchblade 600 was identified as one of the first systems integrated into the initiative. This loitering munition can engage armored targets at ranges exceeding 40 kilometers and fly for approximately 40 minutes, according to its manufacturer. This is not a futuristic experimental drone. It is a weapon that is already known, tested, and adapted to a pressing need.
The Pentagon subsequently added drones such as Anduril’s Ghost-X and Performance Drone Works’ C-100, with capabilities intended primarily for the company level. The goal is not to create a “super-drone.” It is to equip units more broadly with systems that improve reconnaissance, local strike capabilities, persistence, and decision-making capacity.
Replicator 2 then shifted part of its focus toward counter-drone operations. This choice makes sense. The threat does not come solely from China. Drone attacks against bases, ships, and infrastructure are now a central concern. The cost of the problem is staggering: it is absurd to fire a multi-million-dollar missile at a drone that costs a few thousand or tens of thousands of euros. This asymmetry has become one of the great nightmares of Western militaries.
The war in Ukraine has made this reality impossible to ignore. In 2024 and 2025, Kyiv massively increased its drone production. Russia has also industrialized the use of Shahed and locally produced variants. In May 2026, Reuters reported that Russia had launched 1,567 drones and 56 missiles against Ukraine in two days. This figure encapsulates the new normal: saturation is no longer a laboratory hypothesis.
The advantages are real, especially against saturation
The first advantage of Replicator is its relative cost. An expendable system can be deployed where a manned platform would be too valuable. It can scout, decoy, saturate, jam, attack, or force the enemy to reveal its radars. This creates friction for the adversary.
The second advantage is dispersion. A force composed of a few very expensive platforms is powerful, but fragile if it is concentrated in one location. A distributed force poses a different problem. It forces the enemy to detect, sort, track, and neutralize numerous targets. In a high-intensity conflict, this cognitive and technical overload can be just as significant as explosive power.
The third advantage is industrial. Replicator is pushing the Pentagon to work with non-traditional companies, including Anduril, AeroVironment, Performance Drone Works, and other specialized players. This broadens the industrial base. It introduces shorter development cycles. It also forces the major historical prime contractors to react.
The fourth advantage is tactical. Autonomous or semi-autonomous drones allow sensors to be extended beyond conventional lines. They can prepare a strike, confirm a target, serve as communication relays, or divert enemy defenses. In the Indo-Pacific, where distances are vast, this distributed persistence is invaluable.
The fifth advantage is psychological. An army capable of mass-producing expendable systems sends a clear message: it can absorb material losses without immediately losing its combat power. This is exactly what the United States has long accused its adversaries of doing better.
The limitations are as serious as the promises
Replicator has an obvious problem: producing thousands of systems is not enough. They must be integrated. They must be stored. They must be transported. Units must be trained. Data links must be established. Communications must be secured. Electromagnetic interference must be avoided. Batteries, motors, sensors, and software must be maintained.
Drone warfare is less simple than it seems. A cheap drone can become useless if GPS jamming is intense. A loitering munition can lose its value if the video link is cut. A swarm can become vulnerable if its command architecture is too centralized. Autonomy can reduce this risk, but it creates another: trust in the machine.
There is also an issue with munitions and effects. Not all drones can destroy tanks. Not all drones can penetrate air defenses. Not all drones can withstand rain, wind, cold, the sea, or electronic warfare. Real combat quickly eliminates systems that are too fragile.
The timeline is another weakness. The initial goal of eighteen to twenty-four months was deliberately aggressive. It had the merit of breaking with tradition. But speed can create blind spots. The risk is buying systems too quickly that will become obsolete just as fast. This is a constant danger in a field where software, sensors, jammers, and countermeasures are constantly evolving.
Finally, industrial dependence raises a strategic question. The United States wants to produce on a large scale, but many components for small drones still come from Asian supply chains. Batteries, sensors, electric motors, and electronic components are sensitive points. A long war is not won with prototypes alone. It is won with factories, stockpiles, and reliable suppliers.
Controversies Center on Autonomy, Secrecy, and Accountability
Replicator raises three major controversies.
The first concerns lethal autonomy. The Pentagon insists on compliance with U.S. rules and appropriate human oversight. But the line between autonomous assistance, target selection, target tracking, and strike decisions can become blurred in a high-intensity combat environment. As speed increases, so does the temptation to delegate to the machine. This is a military issue, but also a legal and moral one.
The second controversy concerns transparency.
Replicator remains partially classified. This is normal for sensitive capabilities. But it makes it difficult to publicly assess the cost, actual volumes, performance, and failures. For a program presented as a major transformation, this lack of transparency breeds mistrust. It also prevents us from knowing whether Replicator truly changes the acquisition system or merely adds a layer of communication to already planned purchases.
The third controversy concerns the overall cost. A drone may be inexpensive per unit. A drone force, however, can be very expensive. It requires networks, cybersecurity, control stations, electronic warfare capabilities, spare parts inventories, simulators, operators, and maintenance teams. Scale does not come for free; it merely shifts the cost.
We must therefore avoid the easy slogan: Replicator does not replace an F-35 with a low-cost drone. It adds a layer of scale around existing platforms. It is less spectacular, but more credible.
Other countries are moving forward with different models
China is the central reference point. It has a very powerful civilian industrial base in drones, high production capacity, and a military that is actively studying swarms, robotics, and smart warfare. The Chinese model combines mass production, military research, civil-military integration, and a doctrine of saturation. Beijing is not merely seeking to copy the United States. It aims to make certain American platforms too vulnerable or too costly to deploy.
Russia follows a different logic. It compensates for its weaknesses in precision and aviation with waves of Shahed drones, decoys, and missiles. The goal is to wear down the Ukrainian defense. The method is brutal, less refined than the American discourse on autonomy, but effective in its economic logic. A cheap drone can force the use of a much more expensive interceptor.
Ukraine represents the most agile laboratory. Its model is based on rapid adaptation, distributed production, FPV drones, long-range strikes, and innovation under constraint. Kyiv has shown that a country under pressure can transform civilian or semi-industrial drones into a mass military system. This is a lesson that Replicator is clearly absorbing.
Israel has extensive experience with loitering munitions, with systems like Israel Aerospace Industries’ Harop. The Israeli model is older, more integrated, and more precision-oriented. It does not rely solely on quantity. It combines intelligence, autonomous navigation, real-time video, and human control.
Turkey, too, has established an export model with Bayraktar, Akinci, and a highly aggressive drone industry. Its advantage lies elsewhere: producing systems that are sufficiently capable, exportable, visible, and tailored to customers who cannot purchase the most sophisticated American platforms.
Europe remains more hesitant. It has strong manufacturers, but it suffers from fragmented needs, budgets, and doctrines. The war in Ukraine has accelerated the thinking, but Europe has not yet found the political and industrial equivalent of a continental Replicator.
The end of expensive platforms is a misreading
The headline is tempting: Replicator would herald the end of the era of expensive platforms. That is false.
Expensive platforms remain necessary for penetration, command, survivability, long-range strikes, and coordination. A B-21, a nuclear attack submarine, an Aegis destroyer, or an F-35 do not perform the same missions as an expendable drone.
They provide range, power, sensors, connectivity, survivability, and strategic impact.
What is changing is their isolation. A high-value platform can no longer be thought of as an isolated tool. It must become the center or partner of a cloud of distributed systems. It must launch, command, receive, or operate drones. It must survive not only through its stealth or defenses, but through the confusion it imposes on the adversary.
The real future is therefore not “fewer expensive platforms.” It is “fewer expensive platforms on their own.” The winning model will be hybrid: a few high-performance platforms, surrounded by numerous cheaper, replaceable, and connected systems.
This is a revolution that is less neat than a slogan. It requires new budgets, new doctrines, new training, and a more dispassionate acceptance of material losses. It also demands a cultural shift at the Pentagon. Buy fast. Test fast. Discard what doesn’t work. Produce what works. Start over.
The real breakthrough will be industrial as much as military
Replicator will succeed if the Pentagon manages to turn a strategic emergency into an industrial routine. It will fail if it becomes a label applied to a few scattered purchases. The challenge is not just to select drones. It is to create an American capability to produce, update, and replace autonomous systems at the pace of modern combat.
The lesson from Ukraine is clear. In drone warfare, the best system in January may be vulnerable by April. An army that takes five years to fix a software flaw will lose to an adversary that modifies its frequencies, cameras, or algorithms in a matter of weeks.
Replicator therefore marks a turning point. Not the death of major programs. Not the disappearance of the F-35, the B-21, or carrier strike groups. But the end of a Western comfort: the belief that technological superiority can always compensate for numerical inferiority. Faced with China, Russia, and wars of saturation, the Pentagon now wants to counter mass with mass. The decisive question is no longer whether this idea is sound. It is. The question is whether America can still produce quickly, in large numbers, and at a cost compatible with a protracted war.
War Wings Daily is an independant magazine.