Ukraine's best weapon is adaptation

Ukraine has relearned how to build cheap weapons in volume and improve them in days. The United States has the capital Ukraine lacks. Each needs the other.

By David Baldwin and Borys Kremenetskyi

David Baldwin, Major General, US Army (Ret.), served as Adjutant General of California, the state partnered with Ukraine under the National Guard's State Partnership Program since 1993. Borys Kremenetskyi, Major General, Armed Forces of Ukraine (Ret.), served as Ukraine's defense attaché in Washington until March 2026.

One of us spent three decades working alongside Ukraine's armed forces as an officer of the California National Guard. The other represented those armed forces in Washington until this spring. We have watched this war from both ends of the same relationship, and we have come to the same conclusion. Washington treats Ukraine as a showroom: inspect the drones, choose a few, and call that adaptation. The thing worth having is the method that produces them.

Ukraine did not begin the war this way. In 2022 it fought with Javelins and 155-millimeter artillery and asked its partners for manned aircraft. The change came in 2023 and 2024, under fire, and it was a change in how the country builds weapons more than in which weapons it builds.

Engineers at the front

The most impressive thing either of us has seen in this war sits just behind the front line. Ukraine has put engineers and scientists there, working hand in glove with the soldiers who use the equipment. When the Russians change a tactic or a radio frequency, the unit and its engineers change the weapon and test it on the spot. They make adjustments daily and field new versions weekly.

The results can be measured. Through Brave1, the Ukrainian government's defense-technology marketplace, a frontline unit's order reaches it in no more than seven days on average. Across the Pentagon's 104 largest programs, the average time to deliver a new capability is more than 12 years.

The pace sets the terms for everything else. The technology on this battlefield changes every two weeks to a month. There is no sense in building a million drones today to use in two years. What you build has to be open to upgrade every few weeks, or it is obsolete before it ships.

The price of a weapon

In August the US Army asked industry for a new missile to shoot down drones and set a target price of under $150,000 each. Ukraine sets its target for an interceptor at a few thousand dollars. Its interceptor drones cost between $1,000 and $15,000, and by Ukrainian military figures they now account for about a third of the Russian aerial targets destroyed over the country.

We still assume that anything made in the United States has to be very expensive, and that assumption is going to hurt us. Since the Cold War we have bought what the Pentagon calls exquisite weapons: superb, costly, and slow to change. In the Iran campaign this spring, Patriot missiles costing more than $3 million each were fired at drones that cost about one percent as much. No country can sustain that exchange. We have forgotten how to build the Sherman tank: a good-enough weapon, made in quantity.

What not to copy

Ukraine has not cracked the procurement code, and anyone who says it has is taking the wrong lesson. Its system is decentralized and centralized at once. A brigade can buy its own drones and small electronic-warfare equipment. It cannot buy an armored vehicle. Brigades that destroy more targets earn more to spend, and roughly 40 percent of brigades account for about 70 percent of the targets destroyed.

That speed has a price. One brigade equips itself with military radios and the brigade next to it uses cell phones, and the two cannot talk to each other. Every unit runs different equipment, so there are no common spare parts and no deeper maintenance system behind them. The central system that buys major weapons remains slow and bureaucratic. An American force that copies the speed without the discipline will get the fragmentation too.

Some of Ukraine's experience will not apply to American forces at all. NATO doctrine expects air power to shape the battlefield before ground forces commit. In Ukraine neither side controls the air, and the war is fought without that assumption. The US Air Force can do things no other air force can. Even so, those assumptions deserve to be tested against what Ukraine is living through, because an adversary watching this war is planning to deny us the air too.

Why buying is harder than it looks

Ukrainian companies tell us the same three things. The first obstacle is legal, and it starts at home. Ukraine opened a new export regime in July, and three months later no permit had been issued under it. A company that clears Ukraine's rules then faces America's.

The second is intellectual property. Founders want a guaranteed route to market before they share a design, and many small firms have never filed a patent. A buyer will test a system on those terms but will not field it.

The third is capital. To test a system in the United States, a company typically has to supply 10 to 15 units, a serious outlay for a firm living on wartime margins. One of us was in the room last autumn when Ukraine presented its drone proposal at the Pentagon. The message was that Ukraine could build 10 million drones a year and was building seven, because the money to build the rest was not there.

Europe has acted on this. Denmark and its partners now pay for weapons built inside Ukrainian factories, about $6 billion in 2025. In June the European Union disbursed EUR 3.9 billion for drone procurement. Britain has licensed a Ukrainian interceptor and is building it at home. The United States has signed a statement of intent to test Ukrainian drones. The broader agreement, which Kyiv says it is ready to sign, is still unsigned.

What Washington should do

First, put engineers forward. Embed physicists and engineers who understand electronic warfare with operating units, give them the tools to build and test on the spot, and connect them to engineers at home who can turn a field fix into a production change. Then build one place that collects what every unit learns and spreads it across the whole force. Calling an office an innovation office does not make it one.

Second, set the price. Require that the next counter-drone interceptor cost thousands of dollars, and that every unmanned system be built on an open design that can be upgraded every few weeks. Buy in batches and expect to replace them.

Third, fund production and clear the path. Do what Europe has done and put money into Ukrainian production lines, in Ukraine and in joint ventures in the United States. Tell Ukrainian companies plainly what corporate structure makes them eligible to sell to the Pentagon, which licenses they need on each side, and how their designs will be protected. The agreement Kyiv is ready to sign is the place to start.

The technology is the smaller part of what Ukraine has to teach. The larger part is how the country is organized to change it, and the United States is nowhere near being able to do the same. Meanwhile Russia's factories produced about 7 million shells, mortar rounds, and rockets last year, and North Korean workers are being recruited to build drones on Russian lines. Ukraine holds the production method and the United States holds the capital. The longer the two stay separate, the more time our adversaries have to build, integrate, and learn from each other.