When the world's at stake,
go beyond the headlines.

For a few years now, Western observers have breathlessly praised Ukraine’s successes in defense innovation, from AI to drones to decentralization and an ecosystem of defense startups. But all is not well in Ukraine.

Until recently, it was very difficult to publicly question such materials due to the lack of publicly available battlefield statistics. But by 2025, many observers acknowledge that Russia may have surpassed Ukraine in certain areas of innovation adoption.

This is now publicly acknowledged by various experts, including Ukraine’s former top military commander, Michael Kofman, and others. This provides an opportunity for a more open discussion about which parts of Ukraine’s real experience are actually effective and applicable — and which are not — for the United States and Western Europe.

I have been directly involved in shaping the core concepts of the Ukrainian tech ecosystem. For example, I was chairman of the AI Development Committee from 2019 to 2023 and led in the development of the AI roadmap for defense. It is important to understand the context and reasons why the Ukrainian ecosystem developed the way it did under the conditions of war.

There are systemic issues and open questions that should prompt reflection within the Western expert community. Specifically, are Ukrainian models and technological solutions truly applicable to future U.S. competition with China and Russia? Or might they instead limit the potential of the Western defense industry? Is there a need for closer examination of developments in Russia and China since 2022?

My goal here is not to critique or be a Monday morning quarterback, but to take a closer look from different angles at the popularized models based on Ukraine’s experience. Indeed, in many ways (not all) Ukraine, the country I was born in, was doing its best with what it had available to it. But it is important to critically consider, or at least raise questions about, whether Ukraine’s experience of defense innovation offers a truly useful model for Western countries in a future conflict against major powers such as China or Russia.

Decentralization Isn’t All It’s Cracked Up to Be

At the start of the active phase of the war, Ukraine ranked 122nd out of 180 countries in the Corruption Perceptions Index. The standard drone certification procedure took one year. The state budget for 2022 was reduced, while the budget for road construction in 2022 was increased threefold.

With the start of the active phase of the war, the state procurement system was effectively paralyzed. One could argue at length with government officials that the system was not paralyzed — but I personally had to buy helmets, vehicles, etc. for friends during the first week of combat. In the second week, I had to procure underwear and socks for an entire battalion near Kyiv. This batch was delivered via commercial mail — Nova Poshta —to the closest non-occupied branch, and the battalion picked it up a week later.

Viewed holistically, something very unusual occurred: due to the paralysis of the state procurement system, military units were allowed to procure everything they needed themselves. At the same time, various charitable foundations and private donors became active players in military procurement, supplying uniforms, food, and weapons directly to units. To meet the demand, volunteers began creating ad hoc information technology procurement systems when the war was only four months old.

All this led to total de-bureaucratization and partial de-corruption of procurement, producing a very interesting phenomenon: Military units and charitable foundations became more stable and predictable customers than the state. As a result, some technology companies producing drones grew within these ecosystems, reinvesting earnings into development — following the example of Ukrspecsystems in 2014 and 2015, which originally raised funds for drones via PeopleProject and became market leaders by 2021.

Sounds like a success story? Probably, yes — but by 2025, this system revealed major systemic problems: Military units began procuring equipment independently with minimal oversight, and corruption spread down to the unit level as a result.

Such decentralization could make it more difficult for the state to focus resources on promising developments or to ensure integration among military products acquired by different units. I’m not even touching on real costs, potential savings from scale, or interoperability between systems.

In all military conflicts, competition takes place not just on the battlefield but also in logistics systems. And the procurement system that actually developed in Ukraine addressed very different problems than those facing the U.S. in competition with China and Russia. A decentralized system may turn out to be far less effective than centralized procurement systems — and Russia is already demonstrating this on the battlefield in 2025.

Zoo of Tech Platforms Are Not Scalable

Since 2022, Ukraine’s decentralized procurement system sparked a surge in small defense businesses. Thousands of companies began producing drones, components, software, and providing services for the front lines. This movement became massive — it was hard to ignore. It truly bolstered Ukraine’s combat capacity at a critical moment. But in terms of quality and sustainability, it’s increasingly being compared to China’s “Great Leap Forward” in the 1950s, when the Communist Party encouraged citizens to smelt steel in backyard furnaces.

In 2024, the Ministry of Digital Transformation officially proposed that people assemble drones at home. After a wave of criticism, the idea was scrapped, but its public launch became a troubling marker of the decision‑making quality.

By 2025, it became clear that only 20 to 40 percent of first-person view remote-controlled drones actually reached their targets. As a result, the true cost of destroying a single tank was far higher than the advertised $500 price tag per drone. This discrepancy was largely due to Russia’s advanced and rapidly scalable electronic warfare capabilities. On the Ukrainian side, few efforts were made to develop systems capable of evading such interference or anticipating future threats — such as fiber-optic drones or autonomous navigation in GPS-denied environments.

The core issue was a lack of both government expertise and engineering capacity. Developing such systems requires deep knowledge in software, electronics, and systems architecture — yet most newly formed companies were little more than “assembly workshops,” lacking the technical depth to go beyond basic assembly. As a result, Ukraine failed to respond effectively to emerging threats: specifically,  short-range fiber-optic drones and long-range drones with autonomous navigation.

In contrast, Russia has reportedly taken a more centralized approach, relying on a pre-selected group of major defense firms to develop core technologies and platforms. These companies focus on a limited set of systems — such as the Geran / Shahed drones — and refine them methodically by incorporating advanced AI capabilities, improved electronic components, and standardized architectures. As a result, Russia is steadily building a more unified and resilient defense technology stack — better suited for mass production, long-term iteration, and sustained conflict under modern warfare conditions.

The Ukrainian model, on the other hand, produced a zoo of solutions — fragmented, often incompatible systems without standards or architecture. This zoo cannot be scaled. Already inside Ukraine, discussions are underway about what to do: the system is losing competitiveness and is hard to develop further. Yes, some criticism comes from large industry players seeking to push small companies — those working directly with the military — out of the market. But these critics have a systemic point: Small teams rarely achieve the kind of complex systems engineering or high-volume serial production needed for advanced weapons. Ukraine’s wartime experience, unfortunately, has so far borne this out.

To be sure, production decentralization played a crucial role in the early phase of the war. Without it, much would not have been possible. That experience is valuable. But as the conflict wore on and quality demands rose, it became increasingly apparent that more centralized, structured approaches tend to outperform decentralized ones by providing not only results but also predictability.

Is there anything in the Ukrainian experience that might be useful for the West? Yes, but only specific techniques, discoveries, and team flexibility. These are valuable, but Ukraine’s overall decentralized model is not. You cannot scale a zoo of disparate systems in a long war; modern war demands solutions that can be scaled.

Listen to this episode of The Insider featuring author Vitaliy Goncharuk by becoming a War on the Rocks member here

A Strategic Illusion?

As of 2021, Ukraine’s defense sector was in extremely poor condition. The dominant player remained the state-owned conglomerate Ukroboronprom (renamed to Ukrainian Defense Systems), which united over 130 enterprises, most of them founded back in the Soviet era — including Antonov and a number of factories engaged in weapons production and equipment repair for export to Africa and the Middle East.

However, Ukroboronprom’s main task before 2022 was not to produce modern weaponry, but to undergo corporatization and implement corporate governance systems.

The private defense sector remained very limited in both scale and capability. Despite the ongoing armed conflict with Russia since 2014, there had been no emergence of new engineering schools or a significant number of competitive businesses in this sector — apart from a few exceptions.

By early 2022, the only talent pool remotely related to engineering consisted of professionals from the software as a service, financial technology, and entertainment information technology sectors — those who had worked for export pre-war and were part of the startup community that knew how to organize grants, pitches, and small funds. When the war began, they turned to what they knew best: Within a few months after February 24, defense startup events, competitions, and accelerators began to appear.

But there was no systematic research, no analysis of the enemy’s technological development, and no articulation of the actual needs of the front. There was no strategy. Most of the funded solutions turned out to be experiments without context — shots in the dark.

The most publicly visible initiative was Brave1, essentially an events organizer and loose network of mentors funding individual teams. It offered networking and pitch opportunities, but conducted no deep analysis of Russian tech trends, no rigorous scenario planning, and developed no unified architectural design. The funding provided by the platform was largely symbolic given the scale of the challenges: by 2025, Brave1 had disbursed only about $8 million in grants and helped attract around $25 million in private investment — roughly equivalent to a single Series A round for one startup in Silicon Valley.

The format that emerged in Ukraine was a consequence of what resources — time, people, and money — were available under the shock of war. It wasn’t the result of strategic planning; rather, it was a reactive solution.

For low-complexity tasks — where hundreds of simple solutions need to be quickly produced — such a model may be useful. But in long-term systemic competition — China and Russia versus the West — the startup-based model doesn’t scale, because its goal is rapid prototyping and problem-solving, not scalability.

The Ukrainian experience revealed something important: you need infrastructure that allows for rapid development, testing, and iteration of a product within a single loop — specifically, with testing conducted directly on the battlefield. But that’s more a conclusion about the importance of engineering and manufacturing discipline — not a case for adopting the startup approach as the core framework for capability development.

The main challenge isn’t to build such a loop during conflict. The conflict will force it to happen. The challenge is how to make it sustainable in peacetime, when there’s no urgency, but there is competition for talent, attention, and resources.

China, for example, manages this at scale: its top engineers work at firms like DJI that operate commercially and for export. Different sources estimate that around 60 percent of the foreign components in Russian drones and up to 80 percent of those used by Ukrainian manufacturers are sourced from China. This continuous market-driven production keeps the entire system — from design to logistics — in practice and in shape.

No cluster of small startups, no matter how creative, can by itself solve such problems. A startup ecosystem lacks the scale and cannot achieve the procurement volume or the systems integration required for national defense architecture. In short, it’s a different beast. Building a sustainable defense engineering platform is not about grants and meetups; it requires the art of strategic statecraft — long-term investment, coordination, and planning at the national level.

The Limits of Agility

The war in Ukraine has undeniably exposed deep systemic problems in military production across the United States and Western Europe, while also clearly highlighting the technological development trajectories of China and Russia. These two countries are emerging as the primary benchmark players in future non-nuclear conflicts — with a focus on system coherence, scalability, and technological maturity.

In contrast, at the outset of the war, Ukraine had a degraded defense ecosystem.  The state sector was focused on internal reforms and maintaining export contracts inherited from Soviet times, while a private defense sector barely existed. The only model that could be rapidly deployed was a decentralized supply system, based on small manufacturers, low-tech solutions, and grassroots startup community initiatives.

This improvised system did indeed provide significant support during the early years of the war. It mobilized thousands of teams, enabled rapid response, and supported the front line. But its limitations were obvious. A lack of strategic planning, fragmented solutions, an inability to scale, and a weak engineering base resulted in the model reaching its limit by 2024-2025. The result was a zoo of technological platforms — uncoordinated, unintegrated, and often ineffective under conditions of modern electronic warfare.

Despite the dominance of the Ukrainian experience in media and public discourse, its value must be assessed with caution — especially on topics such as procurement decentralization, leveraging startup ecosystems for scalability in defense, the growing role of low-cost remotely controlled drones, and more. It offers lessons in what to do when your defense ecosystem has collapsed, but it does not answer how to build a mature defense model for the long term. What could truly be valuable is a comparative analysis, integrated with an understanding of how long-term engineering, technological processes, and ecosystems are organized in competing countries like China and Russia.

Ukraine’s defense startup ecosystem cannot be seen as a strategic alternative to mature state systems. It falls short when it comes to mass production, architectural compatibility, export viability, and resilience. It is suitable for short-term mobilization — but not for systemic competition.

The United States and its allies may benefit from looking beyond the dynamics of the current conflict and considering the nature of potential wars that could emerge over the next two to three years — conflicts likely to require a fundamentally different scale and level of technological coordination. Meeting those future demands will likely depend on the ability to scale rapidly, concentrate efforts, and invest in long-term engineering platforms. Notably, China and Russia appear to be moving in this direction by streamlining the number of platforms, aligning with industrial partners, and planning across multi-year horizons.

In such a strategic landscape, the decisive factor may not be the sheer number of startups or low-cost drones, but rather a nation’s capacity to develop deeply integrated, scalable, and resilient defense ecosystems. While the Ukrainian experience offers valuable lessons — especially in terms of agility and mobilization — it should be adopted with care. Transplanting that model wholesale, without accounting for the structural differences in Western defense institutions and industrial bases, could actually undermine the effectiveness of Western strategies in the long run.

Vitaliy Goncharuk is an American entrepreneur with Ukrainian roots, known for his work in autonomous navigation and AI. In 2022, his company Augmented Pixels was acquired by Qualcomm.

From 2019 to 2023, Vitaliy served as chairman of the AI Committee of Ukraine, sat on the External Innovation Board of UkrOboronProm (renamed to Ukraine Defense Systems), and advised the Vice Prime Minister of Ukraine for Temporarily Occupied Territories, who later became Minister of Defense.

Image: ArmyInform via Wikimedia Commons