Interview with Sam Rosenberg

CSDS-SWJ STRATEGY DEBRIEFS • 12/2026

Interview with Sam Rosenberg, by Octavian Manea

16.7.2026

Lt. Col. Sam Rosenberg serves as the Concepts and Contingencies Branch Chief in the G5 Plans Division for US Army Europe-Africa, headquartered in Wiesbaden, Germany. In this capacity, he led the development of the Eastern Flank Deterrence Initiative (EFDI), the command’s warfighting concept to enhance NATO’s collective defence. This Strategy Debrief examines some of the key lessons from the war in Ukraine that have shaped the emerging EFDI, a new NATO warfighting concept. Rooted in the perception that ground deterrence in Europe had weakened in the wake of Russia’s full-scale invasion of Ukraine, EFDI advances a form of deterrence by denial built around cost-effective mass and autonomous systems. Its central aim is to move beyond reliance on punishment after an attack and instead create an integrated regional defence architecture capable of deterring aggression from the outset and preventing Russia from rapidly imposing a fait accompli.

How would we define the contemporary Russian way of war? Does it more closely resemble the opening phase of the invasion of Ukraine – a rapid, surprise-driven campaign aimed at achieving a fait accompli – or the protracted, industrial-scale war of attrition that the conflict has since evolved into? More broadly, what enduring characteristics of Russian military thinking and operations should (NATO) planners focus on

Sam Rosenberg: When I look at the national security community, I think what we are seeing now is a really interesting debate between attritional warfare and manoeuvre warfare. To a certain extent, people are trying to understand whether Russia is focused on manoeuvre – on conducting some sort of fait accompli against NATO – or whether it is going to leverage a more traditional form of attritional warfare.

The way I look at this, however, is that I do not necessarily see it as a binary choice. I do not think it is one or the other. I think a lot of it is about sequencing.

I would see Russia’s preferred way of war – its preferred theory of victory – as being closer to manoeuvre or fait accompli. When we look at Georgia in 2008, Crimea in 2014, as well as the opening phase of the war in 2022, the Russians largely followed the same design. They wanted to move fast, consolidate their gains, present an accomplished fact on the ground and then challenge the other side to reverse the facts on the ground. That is how I see their preferred way of fighting and what I think they would want to do against NATO.

What we are seeing now in Ukraine – this kind of protracted, industrial war of attrition – is their fallback option. It is not their preference. In many ways, it may be what they are culturally and industrially built to endure. But I do not necessarily think it is their preferred way of doing business. They are losing more than 1,000 soldiers a day, a pretty staggering number for any country to absorb.

So, from a NATO perspective – from my foxhole as a planner – I do not necessarily want us to plan against the way Russia is fighting now in Ukraine, which is heavily focused on attritional warfare. I want us to plan against the way I think Russia would seek to fight NATO: through a fait accompli, move fast, win quickly, act before the Alliance can make decisions collectively, seize a piece of the Eastern Flank and then hide behind nuclear coercion to freeze the result. That is why the fait accompli scenario is, I think, our driving planning factor and the outcome we are seeking to prevent.

In NATO planning writ large, the change you have seen over the past several years is a shift from an emphasis on deterrence by punishment towards deterrence by denial.

Prior to 2022, I think NATO was very focused on deterring Russia by threatening to punish it severely after the fact – after it had already conducted an incursion. What we are learning from Ukraine is that, if Russia is able to seize ground, reversing those gains becomes extremely costly. We cannot afford to allow Russia to take any ground at all. We have to deny it from the outset.

That is why we are focused on deterrence by denial. And that is where the EFDI comes in. It is a deterrence-by-denial warfighting concept. We want to deny Russia from the outset, not rely solely on punishing them after the fact. We want to prevent the incursion from ever happening in the first place.

What key operational and strategic lessons should be taken from the war in Ukraine? To what extent have such lessons – and the broader trends emerging from the conflict – reshaped thinking about deterrence and defence along NATO’s Eastern Flank?

Sam Rosenberg: There are five main lessons that stand out. The one that jumps out to me the most – and the one that I think has really influenced the development of EFDI – is the idea that it is very difficult to surprise enemy forces on the battlefield. This has sometimes been shorthanded as the “transparent battlefield”. The rise of persistent, pervasive surveillance is lesson number one. It is something that we all have to think hard about. Because drone technology and sensors are now so pervasive on the battlefield that any time you move, you can be seen. Any time you can be seen, you can be targeted and killed. It is therefore very difficult to mass forces and manoeuvre them. The transparent battlefield and the rise of pervasive surveillance are, I think, the first lesson that we all need to think through.

The second lesson that we are drawing from Ukraine – and that has influenced EFDI – is the idea that affordable mass will likely have an advantage over exquisite scarcity. The American way of fighting has traditionally relied on exquisite, highly capable systems, but in relatively small numbers. What Ukraine is showing is that affordable mass currently has a distinct advantage over exquisite scarcity. We are therefore trying to think through new ways of building magazine depth through attritable systems and consumable systems. That does not mean writing off exquisite scarcity altogether. We still need ATACMS and Patriot interceptors, for example. We are not discounting those capabilities. What we are trying to emphasise is an optimal mix of exquisite scarcity and affordable mass.

The third lesson concerns large-scale combat. What we are seeing in Ukraine is that large-scale combat is obviously a military contest – a clash of armies on the battlefield. We all understand that. It is a different way of fighting from the one in which I grew up in the Army, which was shaped by counterinsurgency operations in Iraq and Afghanistan. But large-scale combat is just as much an industrial contest as it is an operational fight. Our defence-industrial base, our commercial counterparts and our teammates in industry are therefore very much part of this effort. Unless we can outpace Russian industry, we will have a hard time on a future battlefield should we find ourselves fighting against Russia. So, lesson number three for me is that large-scale combat is just as much an industrial contest as an operational one.

The fourth lesson that is actively shaping EFDI is the primacy of data and rapid adaptation. What we are seeing in Ukraine is that the digital sensor-to-shooter cycle is compressing at an extraordinary rate. This new reality, we think, demands a secure data backbone that can connect different national systems and allow them to interoperate securely across our 32 nations. Critically, this architecture will ensure data sovereignty. This is a shift from legacy, slow-moving processes that we used in the past to share information. In fact, building the EFDI’s data backbone is our top focus right now in the Baltics and Poland. To make this a reality, though, we have to dramatically shorten our traditional acquisition timelines and move from cycles that used to take years toward software and hardware updates that can be deployed in weeks.

The last lesson that I would offer is that we often assume that, if we attrit or destroy a certain proportion of Russian forces, they will eventually give up and concede. That reflects our Western way of thinking. We are highly casualty conscious. If we were to sustain the levels of casualties that Russia is enduring right now, it is unclear to me whether we would be able to stomach them. One of the lessons we therefore have to consider is how to translate military and tactical gains on the ground into lasting political outcomes when fighting an enemy that does not necessarily view battlefield losses in the same way that we do.

Let us unpack what the EFDI is really about. What problem was it designed to solve and how did the concept emerge? What are its key components and where does implementation stand today?

Sam Rosenberg: Let me explain the problem that gave rise to the EFDI and then talk through how the concept came about.

We have often been asked: why now? Why is NATO pushing a new warfighting concept? After all, we have just gone through a major adaptation process since 2022.

Our answer is that we think there is a serious Russian mass-and-momentum challenge that we will face within approximately two years of a cessation of hostilities in Ukraine. We believe the Russians will be able to modernise, reconstitute their forces and present a credible mass-and-momentum challenge on NATO’s Eastern Flank.

What does the mass-and-momentum challenge mean? I like to think of it in three parts.

First, the Russians are able to mass significant forces on NATO’s borders relatively quickly. In 2022, they assembled approximately 150,000 to 190,000 soldiers, depending on which media reports you read. They were able to mass those forces relatively quickly under the guise of exercises and then move rapidly across Ukraine’s borders. For NATO to do something comparable would obviously take much longer and would be far more difficult. So Russia can mass large forces in a relatively short period of time.

Second, on the NATO side, we all know that there are force-structure challenges within NATO formations. There are also force-posture challenges, in the sense that some of the forces we expected to be available in particular locations may not be present there in the future. So, while Russia has a great deal of equipment and can generate substantial mass, NATO may have a relatively limited forward force structure within two years of a ceasefire in Ukraine.

The third element of the Russian mass-and-momentum challenge is the anti-access/area-denial (A2/AD) problem. Russia has three principal A2/AD complexes on the Eastern Flank. We are laser-focused on these because they will challenge our ability to reinforce our limited forward force structure and complicate our ability to move combat power around Europe in the event of a crisis.

Those three issues, taken together, create a Russian mass-and-momentum challenge within approximately two years of a cessation of hostilities in Ukraine. Russia can generate substantial mass. NATO has a relatively limited forward force structure. And Russia has A2/AD systems that will complicate our ability to reinforce the Eastern Flank and project combat power.

So, what do we do about that?

When General Donahue came in, in December 2024, he recognised that, to a significant extent, we had lost ground deterrence in Europe. We needed a new way of re-establishing it. And it would take both NATO’s Allied Land Command and US Army Europe and Africa to address this challenge in the land domain.

From the moment General Donahue took over in December, we immediately began developing a new warfighting concept for NATO’s land domain. It was a fairly robust effort. It began with a counter-mass-and-momentum wargame in April 2025, where we gathered around the table, simulated a Russian invasion and saw the problem firsthand.

We had participants from across the joint force. It was a US European Command-led event, so the joint components were represented. We began brainstorming new approaches to the problem, drawing on the emerging technologies we were seeing in Ukraine.

From there, we started building the components of EFDI.

First, we developed a concept for taking down A2/AD complexes using persistent sensing and affordable, attritable mass.

Then, in approximately May 2025, General Donahue said: I want to combine this effort to take down A2/AD complexes with an effort to generate greater capacity across the Eastern Flank. How can we do that? How can we create new forms of mass across the Eastern Flank?

He asked us to develop a defence-in-depth concept using uncrewed systems. We therefore incorporated that element into EFDI. Over the summer, he began rolling the concept out to Allies and partners. The concept was unveiled more broadly, and over the following year General Donahue continued travelling around the Alliance, discussing it and taking a series of measures to turn it into a reality on the ground – to move it from concept to implementation.

We are doing that in several ways.

One is experimentation. On the US side, we have experiments that we call Accelerated Capability Events. In each case, we focus on a specific capability, such as ground autonomy or air autonomy. We invite industry partners into our training areas and allow our soldiers to use their equipment. We identify the best-performing systems and then move them rapidly into the procurement process.

On the NATO side, we are doing something similar through Task Force X. NATO holds experimentation events where industry partners bring in their systems, we test their capabilities, identify the strongest performers and move them into acquisition and procurement processes.

Experimentation is one way in which we are making the concept real.

Another is the way we are revamping our training. Recently, we have brought Ukrainian soldiers into training rotations in Germany to challenge our forces and help us think through new ways of fighting. Ukrainian soldiers have fought against our forces during training. We have also embedded Ukrainians within our forces to see how much additional capability we gain by having them on our side and incorporating their new ways of fighting.

All of this is to say that EFDI emerged because, to a certain extent, we had lost ground deterrence in Europe and needed to re-establish it. We will face a Russian mass-and-momentum challenge if we do not start thinking and preparing differently. And we are turning the concept into a reality today through several methods, most notably experimentation and training.

Our Allies and partners are already moving forward. Poland has its East Shield programme and is building bunkers and deploying counter-UAS systems. The Baltic states are doing something similar through the Baltic Defence Line: installing traditional defences and positioning sensors along the Eastern Flank.

So this is no longer merely a concept. It is becoming a reality.

Can we stay for a minute on unpacking the concept for taking down A2/AD complexes using persistent sensing and affordable, attritable mass? This remains fundamentally a cross-theatre problem and essential to modern warfare?

Sam Rosenberg: You are right, it is a cross-theatre problem. For us here, we think of three main tools to help us do this faster than before. First, we are working hard to integrate persistent sensors on the ground and in the air. This means EO (Electro-Optical)/IR (Infrared) and acoustic sensors positioned along the Eastern Flank to identify indications and warnings as early as possible. In the air, we are looking to leverage long-range, long-endurance systems to have a persistent stare on enemy formations.

We think these sorts of ground and air sensors can “tip and cue” a range of traditional exquisite systems and non-traditional long-range, cost-effective fires. On the non-traditional side, we are emphasising launched effects and long-range one-way attack systems to saturate enemy defences.

This would all be connected by the EFDI data backbone to help us decide and operate faster than the adversary.

How is EFDI supposed to function?

Sam Rosenberg: The way this would work on the ground – focusing on defence in depth through uncrewed systems – is that we envisage an area along NATO’s Eastern Flank that would primarily function as an uncrewed zone. It would leverage uncrewed aerial vehicles, ground vehicles and sensors to help us detect and track Russian movements.

Those systems would be tied directly to new, affordable and attritable forms of mass: launched effects, one-way attack drones and bomber drones. Once Russian forces have been identified, we could begin disrupting and degrading them. We would then leverage traditional and non-traditional fires to destroy the forces that remain.

We therefore see this uncrewed zone as consisting of three layers: a detection zone, a disruption zone and a destruction zone.

The detection zone is centred on persistent surveillance in the air and on the ground. It would “tip and cue” new forms of affordable, attritable mass that could be used to disrupt Russian forces.

Those disruptive effects would then pave the way for traditional long-range precision fires, as well as newer and more affordable long-range systems, including longer-range drones and containerised munitions. These capabilities could then be employed to destroy the forces that remain.

As I mentioned earlier, we think that, within two years of a cessation of hostilities in Ukraine, Russia could pose this mass-and-momentum challenge along NATO’s Eastern Flank. We believe that Russia will have more forces available than NATO will have forward-deployed.

The purpose of the uncrewed zone is therefore to reduce Russia’s initially favourable correlation of forces and means – in other words, its numerical advantage over NATO – to a level at which NATO forces can manoeuvre effectively. The aim is to move from a highly unfavourable force ratio to one that is more favourable from NATO’s perspective.

Those three layers would be connected by that EFDI data backbone I mentioned earlier. Essentially, different countries would retain their own sovereign cloud infrastructure, allowing them to store their data securely. At the same time, one country could detect an incoming threat, pass that information to another country and enable that second country to engage the target if necessary.

The aim is to achieve seamless data-sharing across national boundaries while maintaining data sovereignty. That is the general concept we are looking to build along NATO’s Eastern Flank.

How does EFDI fit within NATO’s broader deterrence and defence posture? In particular, what is the relationship between the EFDI and the Alliance’s more traditional force structures such as the forward-deployed multinational brigades on the Eastern Flank and the follow-on reinforcement echelons that would deploy in a crisis or conflict? Is the EFDI primarily intended to complement these forces, enhance their effectiveness or alter the way NATO thinks about forward defence altogether?

Sam Rosenberg: First, I want to emphasise that the NATO battlegroups you mentioned are all part of the Alliance’s new planning construct: the regional plans. All of that still exists and remains the law of the land from our perspective. EFDI is not designed to compete with that construct. It is not designed to invalidate it. It is supplementary and complementary to it. EFDI enables NATO’s regional plans and Forward Land Forces (FLFs) to defend every inch of Alliance territory.

We see EFDI as the warfighting concept that enables NATO’s regional plans. It is the concept that will enable NATO’s forward corps to fight more effectively and efficiently.

The goal of the EFDI is really to preserve human combat power. We can replace drones, we can replace sensors, but it is very difficult to replace a NATO soldier. We therefore want EFDI to do a lot of the heavy lifting against Russian forces: to attrit those forces and shift the force ratios in NATO’s favour, while preserving our human combat power for the tasks that human forces are currently best positioned to perform – manoeuvring and taking ground. At the moment, we do not have uncrewed ground systems that can perform those tasks particularly well. We therefore want to preserve our manned combat power for manoeuvre.

That emphasis on manoeuvre is really important. People often look at EFDI and say: this is simply a defensive concept. But it is a warfighting concept. A central part of it is maintaining the ability to manoeuvre and to hold at risk something that Russia values. The true deterrent value lies in the ability to manoeuvre. We therefore want to preserve human combat power by leveraging uncrewed systems to do much of the heavy lifting.

There is a longstanding argument that NATO’s way of warfare is, to a significant extent, an extension of the American way of warfare – relying heavily on a range of key US strategic enablers. How dependent is the concept on these/such strategic enablers? Could the EFDI function as intended in a scenario where US support was delayed, constrained or focused on another theatre?

Sam Rosenberg: At the conceptual level, NATO’s warfighting model remains heavily dependent on American enablers. From strategic airlift and space-based capabilities to long-range fires, theatre-level ISR, integrated air and missile defence, command and control, intelligence – much of NATO’s warfighting model is still heavily enabled by the United States. To suggest otherwise would be unrealistic.

Obviously, this is changing. We are seeing considerable initiative from our Allies and partners to increase defence spending and assume greater responsibility for security in Europe. That is a very positive development. But the reality is that, after seventy or eighty years, the United States still enables a significant share of NATO’s warfighting effort.

The really positive aspect of EFDI is that it is not only a warfighting concept. It is also a vehicle for burden-sharing. It gives Allies something around which they can orient their increased defence spending. Most Allies have signalled that they are prepared to spend more on defence. EFDI provides a concept they can rally around, focus on and invest resources in.

So, while EFDI is obviously a warfighting concept, we also see it as a vehicle for enabling burden-sharing. The US would not be the principal actor in the defence-in-depth component of EFDI – that uncrewed zone I mentioned earlier. The units responsible for implementing that concept would largely be Finnish, Polish, Latvian, Lithuanian, Estonian, German, Dutch and other frontline nations – not American. In that sense, EFDI enables our Allies and partners to assume primary responsibility for security in Europe.

EFDI is a practical concept for doing exactly that – it gives Europeans the means to take on that responsibility.

To an outside observer, the EFDI appears to share some similarities with the “Hellscape” concept being developed for the First Island Chain in the Indo-Pacific – both seek to deny the massed formations of the adversary the freedom of manoeuvre, leveraging dense networks of sensors, fires and autonomous systems. Do you see the underlying principles and capabilities as transferable/fungible across theatres? To what extent are these concepts manifestations of a broader transformation in warfare rather than theatre-specific solutions?

Sam Rosenberg: One hundred per cent. EFDI was born and developed on NATO’s Eastern Flank, but the ideas underpinning it are applicable globally. We see it as a global warfighting concept that can be applied in any theatre. The intellectual DNA is largely the same. How the concept is applied in a particular theatre may change, but the underlying logic is broadly applicable around the world.

You mentioned Hellscape and the Indo-Pacific. Hellscape is the warfighting concept being developed for the Indo-Pacific. In many ways, EFDI and Hellscape are trying to answer the same question: how do you deter an adversary whose theory of victory is based on a rapid fait accompli and on using large numbers of forces to overwhelm friendly forces? Both EFDI and Hellscape are offering a form of deterrence by denial based on mass and autonomy.

What Hellscape seeks to do in the Taiwan Strait through uncrewed maritime and aerial systems, we are trying to do on land in the Suwałki Gap and the Baltic approaches through layered, land-centric effects. The intellectual DNA is the same, even though the way the concept is applied in each theatre will obviously differ.

We have talked about the Indo-Pacific, but one of the really interesting EFDI stories today concerns the Middle East.

Through EFDI, we had been developing and experimenting with several counter-UAS systems. One of them is MEROPS, a very impressive drone-interceptor system. We had been incorporating it into our posture on NATO’s Eastern Flank in response to Russian drone and balloon incursions in the autumn. Several of these systems had already been deployed in countries including Poland and Romania, and we had seen encouraging results in our experiments.

Then Operation Epic Fury began and we saw that there was a serious challenge in protecting some of our US assets in the Middle East. Because we had already been doing this work in Europe – experimenting with the technology and developing EFDI – we had a solution that could support another theatre. We had the MEROPS system. We were able to say: we have a capability that can help.

The system was deployed within days of the beginning of Operation Epic Fury and used to help protect American assets. It was a genuine example of an idea developed and refined in Europe being deployed globally in support of another priority theatre.

I think that highlights the way we currently see Europe: as a battle laboratory. Europe is the theatre in which sustained, large-scale ground combat operations are taking place on a daily basis. We are drawing lessons from that experience, using them to improve our own capabilities and trying to share those lessons as widely as possible.

Technical standards will be an important piece of this across the alliance. We are now in the early stages of creating standards so that we can ensure we are connecting the best sensors to the best shooters as effectively and efficiently as possible regardless of national borders, all while respecting each nation’s digital sovereignty.

A recurring theme in both the war in Ukraine and the EFDI is the need for “affordable mass”. Is the greatest challenge today technological, operational or industrial?

Sam Rosenberg: I see this primarily as an industrial challenge. We know what to build, we can see it in Ukraine, in our own concepts and we are even working through how to employ these systems. The problem is that we cannot field them at scale or at an affordable price and we cannot yet replace them quickly enough.

So, to answer the question very directly, I think this is primarily an industrial-capacity challenge. How do we enable our industry partners to build at the scale, at the price and at the rate required to make this concept relevant and effective?

That is not to say that the technological or operational dimensions are unimportant. They very much matter.

• On the technological side, there are still gaps that we need to address: electronic-warfare resilience, autonomy authorities and the question of how to counter counter-drone capabilities. These are real technological challenges. But they are solvable engineering and policy problems, not fundamental unknowns.

The real roadblocks, in my view, are industrial and procurement-related. There are several dimensions to the industrial challenge.

• The first is magazine depth: we have seen in Ukraine and now in Iran that our munitions stocks and production rates are not at the level required to sustain operations at the necessary tempo.

Sam Rosenberg: There is a great deal of continuity between these ideas. From a planner’s perspective, we see EFDI, in many ways, as a new version of Air-Land Battle. It is certainly inspired by Air-Land Battle, which sought to extend the battlefield. How do you get after the second and third echelons of Soviet forces before they reach the close fight? We are thinking through many of the same questions today.

__________