# OPEN LETTER: To the Generals Who Still Thought Drone Swarms Were Science Fiction

> To you, senior officers, strategic planners, defense procurement bureaucrats who spent years reading reports on autonomous drone swarms and telling yourselves it is ten years away. Take note: it is now. On Salisbury Plain, in England, Australian, American, and British soldiers ju

- Author: Maxime Marquette
- Publisher: MadMax (https://mad-max.co)
- Section: Open letter
- Published: 2026-06-27
- Canonical URL: https://mad-max.co/en/article/lettre-ouverte-aux-generaux-qui-pensaient-encore-que-les-essaims-de-drones-relev

## Key takeaways

- Three weeks of experiments involved Australian, American and British forces, with swarms scaling from two to six drones and sharing AI models across national platforms.
- The main achievement cited is interoperability: drones running different software from different companies shared data in real time under common standards.
- The article points to Ukraine as an accelerating testing ground and to the US Replicator program, and reports that Turkey carried out a live armed swarm strike in January 2026.
- The author raises command questions about human oversight of autonomous strikes and calls for clear ethical and legal rules.
- Defensive saturation and countering swarms, including China's drone programs and Taiwan deterrence, are presented as central strategic challenges.

## Introduction: Salisbury Plain, June 2026 — Drones Fly in Packs, War Changes

### An experiment on the English plain that redefines doctrine

To you, senior officers, strategic planners, defense procurement bureaucrats who spent years reading reports on autonomous drone swarms and telling yourselves it is ten years away. Take note: it is now. On **Salisbury Plain**, in **England**, Australian, American, and British soldiers just completed three weeks of intensive experimentation as part of the **Army Warfighting Experiment 2026**. The results are no longer theoretical. Autonomous drone swarms flew over English woodland, identified targets in real time, shared data between platforms from three different nations, and refined their artificial intelligence models mission after mission.

The exercise, conducted under the framework of **AUKUS Pillar II** — the advanced capabilities strand of the partnership between **Australia**, the **United Kingdom**, and the **United States** — was not a public relations exercise. It was operational experimentation under « **demanding conditions** » with « **near-constant rain** ». Soldiers, engineers, and scientists from the three nations worked together on a central problem: how to ensure that drones of different nationalities, running different software, communicate in near-real time and operate as a coherent swarm in a contested environment? The answer, after three weeks, is: it is possible, and it is improving rapidly.

### From two to six drones, from demonstration to operation

What stands out in the exercise account published by **Australian Defence** on **June 22, 2026** is the progression across the three weeks. The activity began with swarms of **two drones** and progressively scaled up to swarms of **six aircraft**. That is not a large number — discussions about mass swarms evoke hundreds or even thousands of units. But the progression illustrates the fundamental principle of this technology: it is a system whose sophistication scales with experience and iteration.

The drones operated over assigned areas of interest with predefined mission profiles. They used swarm formation software to divide the sector, sweep it, and send images to the end-user's device. The AI model was continuously refined through repeated identification and confirmation cycles — the more a drone recognized targets in the English woodland, the better it became. A **shared command hub** enabled the near-real-time exchange of mission data and sensor information between AUKUS partners. For the first time, distinct national platforms shared AI models under common interoperability standards.

*To those who still smirk at the idea of drone swarms as a real weapon of war: Colonel Jake Penley, director of the Australian Army's Robotic and Autonomous Systems implementation office, put the reality plainly — « Autonomous systems are no longer experimental, they're becoming central to how land forces operate, particularly in contested environments. » This is not science fiction. This is doctrine in formation.*

## What the AUKUS Exercise Really Proved

### Interoperability: the real technical challenge solved

The greatest technical achievement of the **Army Warfighting Experiment 2026** is not that drones flew in a swarm — military experimenters have been doing that for years. The real achievement is **national interoperability**: Australian, British, and American drones running different software, developed by different companies to different standards, successfully shared data in real time and operated as a unified swarm. That is the problem that had blocked every previous exercise.

Engineer **Hing-Wah Kwok** from the **Australian Defence Science and Technology Group (DSTG)** articulated the breakthrough precisely: « **Being able to train and retrain new models in the field and share them rapidly across AUKUS partners under common standards is a true force multiplier.** » That sentence deserves reflection. It means an AI model trained on a specific theater of operations — say, recognizing Russian military vehicles in a Ukrainian forest — can be shared and deployed on a partner nation's platforms within hours. That is a revolutionary capability.

### AI-assisted recognition in woodland: a lesson for Ukraine

One of the most complex scenarios tested during the exercise was **target recognition in dense woodland**. Identifying military vehicles or infantry positions under tree cover is one of the classic problems of land warfare — artificial vision systems struggle to distinguish thermal and optical signatures in vegetation. The AUKUS exercise proved that swarming drones, combined with real-time-refined AI models, can progressively overcome this challenge. The improvement in the ability to « **correctly identify targets in difficult wooded terrain** » is explicitly cited as an exercise result.

For **Ukraine**, this lesson is directly applicable. The wooded areas of the **Donbas** and the Russian defensive lines concealed in vegetation are among the most challenging environments for current Ukrainian drones. Swarms of drones capable of recognizing and designating targets in those environments would represent a significant tactical advance. The fact that the exercise was conducted in English woodland in the rain — conditions comparable to Ukrainian operational zones — is probably not a coincidence.

*I want to be honest about the limits of my analysis: I am not an autonomous systems engineer and I cannot technically assess whether Salisbury's results can be transposed to a real operational theater under fire and electronic countermeasures. What I can observe is the systematic, documented progression of a capability that, three years ago, was still a promise. It is now a real military exercise result.*

## The Transition from Science Fiction to Operation: What Changed?

### Ukraine as an accelerating laboratory

The transition of drone swarms from operational concept to near-deployable system has accelerated spectacularly thanks to **Ukraine**. Ukrainian forces conducted, according to the **Modern War Institute** at West Point, what appears to be the first fully unmanned combined arms operation in military history, in December **2024** near **Lyptsi**, involving unmanned ground vehicles and FPV drones with no infantry participation. While that is not a « swarm » in the strict technical sense — many platforms were individually controlled — it is a demonstration that autonomous and semi-autonomous platforms can execute complex tactical tasks without soldiers on the ground.

In parallel, companies like **Helsing AI** are delivering AI drones to Ukraine, and American firms like **Anduril** tested their autonomous drones in Ukraine before commercializing them. Ukraine is the only operational theater in the world where autonomous systems are being tested under conditions of intensive real combat, against an adversary capable of sophisticated electronic countermeasures. The data gathered in that real-time laboratory accelerates the development of systems like those tested at Salisbury Plain in a way that no simulated exercise could match.

### The Replicator program: the United States appropriates the Ukrainian lesson

The Pentagon's **Replicator** program is the institutional manifestation of that awakening. Launched in **2023**, it aims to deploy thousands of autonomous platforms rapidly, explicitly drawing lessons from the Ukrainian experience. Its objective is to counter China's quantitative advantage in a hypothetical Pacific confrontation by saturating enemy defenses with swarms of low-cost systems. That is the direct application of swarm doctrine to strategic competition with Beijing — and that is precisely why AUKUS exercises take place in this multi-partner format.

**Turkey** also conducted in **January 2026** the first real strike of an armed drone swarm, with **20 autonomous KARGU** loitering munitions coordinated, striking targets simultaneously in a real-fire environment with live warheads. That is the first real-conditions validation — by a nation other than the United States or China — of an autonomous swarm doctrine with kinetic capability. The world is changing, fast.

*There is an irony in the fact that Turkey — a country with a complicated relationship with its NATO allies — is among the first to deploy armed autonomous drone swarms. While the major democracies debate the ethical implications of lethal autonomous weapons, pragmatic states are moving forward. That is not an argument for ignoring the ethical questions — it is an argument for resolving them quickly.*

## The Challenges of Defensive Saturation

### The command question: who decides when an AI recognizes a target?

One of the most complex problems raised by autonomous swarms is that of **command and control**. In the Salisbury exercise, human operators « **assigned areas of interest and mission profiles** ». Drones then executed their tasks autonomously. But at what point can a drone decide to engage a target without explicit human confirmation? The current rule in most AUKUS systems is « human-on-the-loop »: a human supervises and can interrupt, but is not necessarily in the decision loop for every shot.

That is a crucial doctrinal point. The larger the number of drones in the swarm, the less individual human control is practicable. At **six drones**, one operator can supervise. At **600**, it is impossible. Armies that deploy large-scale swarms will have to solve the problem of distributed command — and that resolution will have significant ethical and legal implications for the distinction between combatants and non-combatants, for accountability over strikes, and for international laws of armed conflict.

### Defensive saturation: the central strategic challenge

The concept of defensive saturation is at the heart of swarm doctrine. The idea is simple: send simultaneously enough platforms to exceed the adversary's interception capacity. **Ukraine** applied it against Russia with hundreds of drones over Moscow. Turkey demonstrated it with 20 KARGU. AUKUS is experimenting with six drones in English woodland. The central question is: **what is the saturation threshold** of an advanced air defense system? No one knows with certainty — and that is precisely why exercises like Salisbury are needed to build real operational doctrine.

Defense against swarms is also under development. Systems like the Ukrainian **SkyFall P-1 SunLong** — an AI-guided interceptor to engage incoming drones — represent the emerging countermeasure. The swarm war will also be an anti-swarm war. And the nations that advance fastest on both dimensions — offense and defense — will hold a significant strategic advantage in future conflicts.

*What impresses me about the Salisbury exercise is less what was accomplished than the speed at which it progressed. In 2023, the first AUKUS exercise of this type involved two drones and a proof of concept. In 2026, it is six drones in difficult weather conditions with real interoperability between three nations. In ten years, what will we be talking about? I do not know. But the trajectory is clear, and it radically changes what land warfare will mean.*

## Swarm Doctrine in Practice: What the AUKUS Exercise Really Showed

### Salisbury Plain, June 2026: six drones in coordinated formation

On **June 22, 2026**, on **Salisbury Plain** in **England**, Australian, British, and American teams conducted the first concrete combat exercise testing drone swarm tactics under the **AUKUS** partnership. This was not a simulation — it was a real-terrain demonstration involving **six autonomous drones** coordinating their reconnaissance and simulated engagement actions. According to Colonel **Penley**, who oversaw the tests on behalf of the Australian Army, this exercise demonstrated **command and control interoperability** between the three nations' systems — a first that opens real operational possibilities.

What the exercise concretely proved: drones belonging to different national systems can operate under common coordination, share data in real time, and collectively adapt to an evolving ground situation. That is the functional definition of a swarm. Six drones is modest. But the logic that allows six to function together can in theory apply to sixty or six hundred — provided command systems and bandwidth keep up. The limit is no longer technological. It is doctrinal and logistical.

### From exercise to operation: the conditions for scaling up

The transition from a **six-drone** exercise swarm to an operational swarm of several hundred in real combat is not automatic. Several conditions must be met. First, **command and control systems** must be able to handle the complexity: coordinating hundreds of autonomous entities in motion, with changing objectives and sometimes degraded communications, requires algorithms of considerable sophistication. Second, **radio bandwidth** must be sufficient in an electronic warfare environment where the adversary will attempt to jam communications. Third, the logistics chain — loading, deployment, recovery or replacement of drones — must be designed to operate at scale.

These challenges are known to military planners. They are not insurmountable. The war in **Ukraine** showed that far lower levels of coordination than those in the **AUKUS** exercise were already producing significant strategic results. The **Salisbury Plain** exercise therefore represents a genuine qualitative leap — an advance toward the operational sophistication that Western armies are seeking to achieve before their adversaries.

*Six drones do not yet make an army. But six drones coordinated by systems from three different nations are the beginning of a military architecture that did not exist last year. I say to all the skeptics: come back in five years, when it will be six hundred. Military history is full of technologies mocked until the day they changed the battlefield.*

## Real Precedents: KARGU in Libya and the Lessons of the First Lethal Swarm

### January 2026 in Libya: the first documented use of an autonomous swarm

Before talking about the future, let us look at the recent past. In **2020-2021**, during the conflict in **Libya**, **KARGU** drones — Turkish-made, **Kargu-2** model — were used in what analysts consider one of the first documented uses of autonomous loitering munitions in real combat conditions. These munitions targeted fighters autonomously, without real-time communication with a human operator. The event was confirmed by a **UN** expert report. That precedent is fundamental for understanding what the **AUKUS** swarm represents in the evolution of autonomous weapons.

This sequence is not anecdotal. It marks the transition from a laboratory concept to a battlefield reality. **KARGU** drones are relatively accessible systems. If limited-resource actors can deploy them in semi-autonomous mode, the world's most advanced armies — **China**, **Russia**, the **United States** — have exponentially superior capabilities. The **AUKUS** exercise at **Salisbury Plain** is not a futuristic experiment: it responds to a reality already present on the world's battlefields.

### The ethical and legal implications of autonomous swarms

The use of autonomous drones in combat raises ethical and legal questions that the **AUKUS** agreement's signatory governments can no longer ignore. Who is responsible when an autonomous swarm kills a civilian by mistake? The algorithm's programmer? The commander who launched the operation? The state that deployed the system? International humanitarian law requires that a human being retain « meaningful control » over lethal engagement decisions. But what does that mean in a swarm of **500 drones** operating in milliseconds? **Colonel Penley** himself acknowledged that military AI ethics questions must be addressed in parallel with capability development.

This is not an academic debate. It is a regulatory emergency. The nations that develop these systems first have a responsibility to set clear rules of engagement before use outpaces doctrine. The **AUKUS** exercise is an opportunity to establish those rules at a multilateral level — an international norm on the autonomy of weapons systems, before less scrupulous actors normalize uncontrolled use.

*I am both fascinated and troubled by drone swarms. Fascinated because the technology is real and its military effectiveness is potentially transformative. Troubled because we are building systems capable of killing autonomously, without the ethical and legal rules to govern them being in place. That is not excessive caution: it is the lesson history teaches us about every radical military technology.*

## The Challenges of Defensive Saturation and Possible Responses

### How do you defend against a swarm?

The military question posed by drone swarms is as daunting for defense as for offense. Against a single ballistic missile, a **THAAD** or **PAC-3** system can be effective. Against a swarm of **200 drones** arriving simultaneously from different directions, with variable trajectories and minimal radar signatures, conventional defensive systems show their limits. The cost of interception by **SAM missiles** can be exponentially higher than the cost of the attacking drones. That is an economic asymmetry that **NATO**'s adversaries deliberately exploit.

Defensive responses under development include: high-energy lasers (capable of engaging targets at the speed of light at very low unit cost), electromagnetic railguns, autonomous counter-drones (defensive swarms against offensive swarms), and advanced electronic jamming systems. But these technologies are still in the development phase for large-scale deployment. In the interim, the capability gap between what offensive swarms can do and what defenses can counter is real — and **AUKUS** exercises aim to close that gap as quickly as possible.

### Defensive saturation as strategy: Ukraine as laboratory

**Ukraine** is the world's laboratory for defensive saturation. Its forces have launched massive waves of drones — **190 to 200 aircraft** in a single night toward **Moscow** on **June 18, 2026** — precisely to saturate Russian air defense layers and allow some drones to get through. **Moscow** intercepted **194 drones** that evening, according to Mayor **Sobyanin** — but several still reached the **Kapotnya** refinery. That is exactly the saturation logic: even with a high interception rate, if you send enough units, some get through.

The **AUKUS** exercise fits into that thinking. The partners are developing offensive saturation capability — to be able to use it, but also to understand how to defend against it. The best way to design a defense against a swarm is to build a swarm. That dual reasoning — offensive and defensive — is at the heart of the strategic value of the **AUKUS** drone program.

*The drone swarm race reminds me of the submarine race in the first half of the twentieth century. First mocked, then credible, then dominant — and finally countered by defensive technologies of equivalent caliber. Military history repeats itself. What changes is the speed at which cycles close. And in **2026**, that cycle is faster than ever.*

## What AUKUS Means for Deterrence in the Indo-Pacific Against China

### China observes and accelerates its own drone programs

The **Salisbury Plain** exercise was not designed in a vacuum. It responds to a reality: **China** already has advanced drone swarm programs. The **CH-901s**, the systems developed by **CASIC** (China Aerospace Science and Industry Corporation), and the demonstrations of **1,000-drone** swarms during official Chinese spectacles are not mere show. They demonstrate industrial and algorithmic mastery that **AUKUS** takes very seriously. The technological competition in the **Indo-Pacific** is also being fought on that terrain.

The **AUKUS** response is deliberately multinational: by combining Australian industrial capabilities, British operational experience, and American technological leadership, the three partners create a collective advantage that none could achieve alone. That logic of capability pooling is exactly what military decision-makers in **Canberra**, **London**, and **Washington** wanted to demonstrate in June **2026**. And it is directly relevant to stability in the **Indo-Pacific**.

### The implications for Taiwan and deterrence

The question of **Taiwan** is implicit in every **AUKUS** military exercise. Military planners thinking through a Taiwan invasion scenario know that **China** would massively use drones to saturate Taiwanese, American, and allied defenses during an initial phase. The ability to counter such a swarm — and to use one's own offensive swarms to strike Chinese military targets — is a central element of any credible deterrence strategy in the **Taiwan Strait**.

The **Salisbury Plain** exercise is therefore, indirectly, a deterrence statement addressed to **Beijing**: **AUKUS** is developing real, not theoretical, swarm capabilities. Whether that actually tempers Chinese military calculations is another question. But deterrence always begins by demonstrating a capability. The exercise of **June 22, 2026** is precisely that demonstration.

*I do not know whether **AUKUS** will actually deter **China** from a military adventure in **Taiwan**. Nobody does. But I know that an alliance that does not demonstrate its capabilities is less credible than one that does. The **Salisbury Plain** exercise sends a real signal. And in geopolitics, real signals matter more than statements of intent.*

## Conclusion: Act Now, on Doctrine and Ethics

### What democracies must do

The AUKUS exercise of June **2026** is a clear signal: autonomous drone swarms are leaving the laboratory to enter operational military doctrine. Western democracies — which currently hold the technological advantage in this domain — have a window of opportunity to shape both capabilities and rules of engagement before their adversaries deploy similar systems with zero ethical constraints. That window is narrow. Colonel Penley said it: autonomous systems are entering the core of land operations. Action on doctrine, training, interoperability, and rules of engagement must happen now.

### China is watching, and taking notes

**China** has developed its own drone swarms and tested them at scale in exercises. The technology is proliferating. Every week that passes without clear doctrine on the employment of autonomous swarms in the law of armed conflict is a week during which the technological advantage of democracies erodes. Salisbury Plain is not a laboratory exercise — it is the prefiguration of a military revolution. And like all military revolutions, it advantages those who prepare for it earliest and most seriously.

*This open letter closes with a direct appeal: stop treating autonomous drone swarms as a future subject. It is a present subject. The AUKUS exercise at Salisbury Plain proves it. Ukraine has demonstrated it on the ground for months. Turkey just acted on it with live warheads. The question is no longer « when? » but « are you ready? »*

**By Maxime Marquette, columnist**

## Columnist's transparency note

### Editorial stance

I am not a journalist but a **columnist and analyst**. My expertise lies in observing and analysing the geopolitical, economic and strategic dynamics that shape our world. My work is to take political strategies apart, to understand global economic movements, to put the decisions of international actors in context and to offer analytical perspectives on the changes that are redefining our societies.

I do not claim the cold objectivity of traditional journalism, which stops at reporting facts. **I claim analytical clarity**, rigorous interpretation and a deep understanding of the complex issues that concern us all. My role is to give meaning to the facts, to place them in their historical and strategic context, and to offer a critical reading of events.

### Method and sources

This text keeps the basic distinction between **verified facts and interpretive analysis**. The rule never changes: a factual claim is published only when a verifiable source supports it, and the sources this article actually used are listed under **Sources**, never here.

**Primary source families** used by the publication, when the story has them: official statements from governments and international institutions, public statements by political leaders, reports from intergovernmental organisations, dispatches from recognised international news agencies.

**Secondary source families**: specialised publications, internationally recognised news media, analyses from established research institutions, reports from industry organisations.

When an article cites statistical, economic or geopolitical data, it comes from the institutions that produce it (intergovernmental organisations, central banks, national statistics offices), and the exact institution is named under **Sources**.

### Nature of the analysis

The analyses, interpretations and perspectives in the analytical sections of this article are a **critical, contextual synthesis** based on the information available, the trends observed and the expert comments cited in the sources consulted.

My role is to interpret these facts, to set them in the context of today's geopolitical and economic dynamics, and to give them a coherent meaning within the larger story of the changes shaping our time. These analyses reflect expertise built through continuous observation of international affairs and an understanding of the strategic mechanisms that drive global actors.

This article describes a documented state of affairs on its publication date, not a prediction: later developments may shift its outlook. No update is promised in advance; when an article is corrected or completed, the change is dated in the text.

## Sources

### Primary sources

- [Australian Department of Defence — AUKUS Partners Test Drone Swarming Tactics — June 22, 2026](https://www.defence.gov.au/news-events/news/2026-06-22/aukus-partners-test-drone-swarming-tactics)

- UK Defence Journal — AUKUS Swarm Drone Tactics Tested at Major Army Warfare Experiment — June 2026

- Militarnyi — Drone warfare context Ukraine — June 2026

### Secondary sources

- Defense News — AUKUS trio conducts drone swarm tests in UK — June 22, 2026

- [Foreign Policy — Autonomous weapons regulation debates — June 2026](https://foreignpolicy.com/)

- [Modern War Institute — Battlefield Drones and the Autonomous Arms Race — January 2025](https://mwi.westpoint.edu/battlefield-drones-and-the-accelerating-autonomous-arms-race-in-ukraine/)

-

---

Source: Maxime Marquette, MadMax. When citing, name the author and link https://mad-max.co/en/article/lettre-ouverte-aux-generaux-qui-pensaient-encore-que-les-essaims-de-drones-relev