PoliticsยทHudson Institute
Policy brief ยท Not peer-reviewed

Ukraine’s Drone War Has Compressed Combat to Machine Speed

Drones now account for most casualties on both sides, hold ground with fewer than 100 soldiers per mile, and run on Chinese components. A new policy memo argues the West is reading the lesson too slowly.

What the Study Found

  • Drones now cause an estimated 70 to 80 percent of all battlefield casualties on both sides of the war.
  • Ukraine’s improvised sea drones, buildable in a garage, have destroyed or damaged roughly 30% of Russia’s Black Sea Fleet.
  • Russia’s $20,000 attack drones force Ukraine to burn Western interceptors costing $450,000 to $3.7 million apiece.
  • Western drone production runs on Chinese parts: Beijing controls 80 to 90 percent of the global drone market.

Sometime in April 2026, Ukrainian formations took a Russian position and, according to the accounts that reached analysts afterwards, not one Ukrainian infantryman was exposed to the ordinary business of assaulting a trench. Aerial drones and ground robots did the whole job. No smoke-wreathed sprint across open ground, no casualty evacuation under fire. Just machines, and somewhere further back, people watching screens.

That single reported engagement is the sort of thing that gets quoted in defense circles for years, and it sits near the center of a new policy memo from Can Kasapoglu, a non-resident senior fellow at the Hudson Institute, published on 2 July. His argument is not that robots have taken over. It is stranger and, if he’s right, more worrying for Western militaries: the war has relocated the tempo of combat somewhere outside the human planning rhythm, and NATO has not caught up.

Kasapoglu borrows a term coined back in 2017, well before the full-scale invasion, for what he thinks is happening. Hyperwar. The idea is that artificial intelligence, cheap autonomous platforms, dense sensor coverage and software together collapse the old observe-orient-decide-act loop until it runs at machine speed rather than staff-officer speed. Whoever can process information, command forces and pull a trigger faster wins, more or less regardless of who has the bigger army.

Substack Sign-up form screenshot

The numbers he assembles are, at minimum, arresting. Estimates for 2025 put unmanned aerial systems behind 70 to 80 percent of personnel casualties on both sides. In 2024, FPV strikes and drone-dropped munitions accounted for roughly 40 percent of combat-damaged vehicles; by early 2025 that figure had reportedly climbed to 60 or 70 percent. A de facto no-man’s-land now runs about six miles either side of the front, buzzing continuously, lethal to tanks and supply trucks and small infantry teams alike.

Thin Lines and Cheap Mass

What that grey zone does to force structure is where the memo gets interesting. Ukrainian fighting positions sit roughly 50 metres apart, held by fire teams of two or three. Do the arithmetic and you get something like 64 to 96 soldiers holding a mile of forward line. Set that against the Napoleonic norm of 20,000 combat troops per mile, or the 3,500 per mile Germany managed on the Western Front in 1915, and you can see the trajectory. Depth substitutes for density. Sensors and drones kill what the infantry never sees.

Kyiv has leaned into this hard. Ukraine’s Unmanned Systems Forces are now reckoned at around 20,000 soldiers; every maneuver brigade fields at least one dedicated drone company or battalion. Former Defence Minister Mykhailo Fedorov has said he wanted every last bit of frontline logistics handed to ground robots, and President Volodymyr Zelenskyy has announced 50,000 unmanned ground vehicles arriving during 2026. In contested places like Pokrovsk, the report says, ground robots already account for 90 percent of military activity.

None of which is cost-free, and the memo acknowledges that Ukraine’s drone dependence is partly a symptom. Drones fill holes where artillery, armour, airpower and manpower ought to be. The recruitment contract aimed at 18 to 24 year olds has not produced the volunteers Kyiv hoped for. Mobilization problems, infantry losses and soldiers drifting away without leave keep widening the gap between elite brigades and weak ones. So the robots are not a flourish of technological confidence, exactly. They’re what you build when you can’t build anything else.

Meanwhile the other side has industrialized. Russia’s Rubicon unit, established in August 2024, has grown from about 1,450 personnel in 2025 to roughly 5,000 by spring 2026, with authorization to reach 9,000. And in May 2026 Russia threw more than 8,000 Shahed-Geran drones and decoys at Ukrainian cities in a single month, double the previous May. The report also relays what Ukrainian interlocutors described to Hudson field tours as “human safaris” behind the lines, Russian drone pilots trained on Ukrainian civilians in Kherson as live targets. That claim is grave, and it rests on the memo’s own field reporting rather than on independent documentation.

What NATO Found Out at Hedgehog

The most uncomfortable material concerns the alliance itself. At Hedgehog 2025 in Estonia, more than 16,000 troops from a dozen NATO countries trained alongside Ukrainian drone specialists pulled temporarily off the front. In roughly half a day, by accounts of the exercise, a Ukrainian team of about ten soldiers notionally destroyed 17 armoured vehicles and conducted 30 further strikes. The allied force, Kasapoglu writes, moved as though the battlefield were still partly opaque. The Ukrainians treated it as transparent. Later that year, off the Portuguese coast, Ukraine’s navy led the opposing force in a NATO maritime exercise for the first time; Ukrainian participants said afterwards that all five scenarios ended in their favour, including one in which an allied frigate took enough simulated hits to have been sunk for real.

Then there is the supply chain, which nobody in Washington or Brussels much wants to discuss. Beijing controls somewhere between 80 and 90 percent of the global drone market, plus about 90 percent of neodymium-iron-boron magnet production, two-thirds of lithium, over 70 percent of graphite anode material. Analysts quoted in the memo call drones “flying smartphones”, which is glib and also basically accurate. Western-componented drones cost considerably more than Chinese-componented ones. Ukraine’s own revolution in cheap attritable mass runs on Chinese parts.

This is a policy memo from a think tank whose stated purpose is promoting American leadership, written by one author, not peer reviewed, and it argues for a conclusion its institution favours: faster procurement, deeper NATO-Kyiv industrial integration, formal channels for Ukrainian combat data. Several load-bearing figures are single-sourced or hedged as reported. Kasapoglu himself cites a Modern War Institute warning against reading Ukraine wrong, and concedes the limits: Russia has built a formidable counter-drone ecosystem of jammers, netting and short-range air defence, and drones still can’t replace artillery, armour or long-range fires. Buy quantity instead of firepower, he warns, and NATO would end up worse off.

Still, the data questions remain. Ukraine has accumulated an annotated corpus of real battlefield imagery and sensor feeds that nobody else on earth possesses, updated continuously as Russian crews bolt cages and camouflage onto their vehicles and Ukrainian recognition models go stale. In early 2026 Kyiv agreed to share it with strategic partners for training AI systems. Whatever else the next decade of Western defence policy turns on, some of it will turn on who gets to learn from four years of somebody else’s war.

  • Study type:ย Policy memo (grey literature); analytical assessment drawing on open-source defense intelligence. Not peer-reviewed; published by a think tank, not a scholarly press
  • Author / affiliation:ย Can KasapoฤŸlu, non-resident senior fellow, Hudson Institute; specializes in open-source defense intelligence and emerging defense technologies
  • Publisher / date:ย Hudson Institute, Washington, DC; July 2026
  • Analytical approach:ย Qualitative synthesis of battlefield reporting, NATO exercise accounts, defense-industry data, and Hudson Institute field tours to wartime Ukraine, framed against historical force-to-space ratios and Lanchester attrition modeling
  • Evidence base:ย Open-source only. No formal dataset, sampling frame, or replication materials. Key figures are attributed to Ukrainian officials, Defense Intelligence of Ukraine, unnamed field reports, and “expert writings” rather than to primary documents
  • Geographic and temporal scope:ย Russia-Ukraine war and NATO’s eastern flank, chiefly 2022 through mid-2026, with historical comparisons from the Napoleonic Wars onward
  • Funding / conflicts of interest:ย No funding statement or conflict-of-interest disclosure accompanies the memo. Hudson Institute does not itemize donors in the document; supporters are listed only in its 2024 annual report
  • Data availability:ย No underlying data, methodology appendix, or citation list is included in the document as supplied
  • Main limitation:ย The memo’s core casualty, cost, and force-ratio figures are estimates from a wartime information environment in which both belligerents have incentives to shape reporting, and several are given without traceable sourcing; the analysis is also explicitly prescriptive toward NATO policy rather than neutral description

Reference

KasapoฤŸlu, C. (2026).ย Ukraine’s drone war: The rise of machine-speed adaptive hyperwarย [Policy memo]. Hudson Institute.ย https://www.hudson.org/technology/ukraines-drone-war-rise-machine-speed-adaptive-hyperwar-can-kasapoglu


Frequently Asked Questions

What does “hyperwar” actually mean?

Hyperwar means warfare in which artificial intelligence, autonomous systems, sensors and software compress the observe-orient-decide-act decision cycle down to machine speed. The term was coined in a 2017 US Naval Institute article, years before Russia’s full-scale invasion of Ukraine. The argument is that advantage shifts to whichever side can process information and act fastest, rather than to whichever side is larger.

Are drones really causing most of the casualties in Ukraine?

Drones are estimated to cause most of the casualties in Ukraine, with figures for 2025 attributing 70 to 80 percent of killed and wounded on both sides to unmanned aerial systems. These are estimates rather than verified counts, and they come from analysis rather than official casualty records. The same trend shows in equipment losses, where drone strikes reportedly rose from about 40 percent of damaged vehicles in 2024 to 60 or 70 percent by early 2025.

Why does China matter to a war between Russia and Ukraine?

China matters to the war because it dominates the supply chains that cheap drones depend on, controlling an estimated 80 to 90 percent of the global drone market along with the bulk of production in neodymium magnets, lithium and graphite anode material. Ukrainian and Western drones alike rely on Chinese components, which are substantially cheaper than American or European equivalents. That dependence is why analysts treat a drone-heavy NATO force posture as a potential wartime vulnerability rather than a straightforward strength.

Why should anyone treat a think tank report differently from a scientific study?

A think tank report should be treated differently from a scientific study because it is not peer reviewed and is usually written to support a policy position rather than to test a hypothesis. This particular memo comes from the Hudson Institute, which states its purpose as promoting American leadership, and it argues for faster Western procurement and closer NATO integration with Ukraine. The underlying facts may well be sound, but the framing is an argument, and several of its figures are single-sourced.

Does this mean infantry are obsolete?

Infantry are not obsolete, and the report is explicit on this point. Drones have made maneuver harder, more visible and more dependent on counter-drone measures, but they cannot yet replace artillery, armour, airpower or long-range fires. Ukraine’s heavy reliance on unmanned systems partly reflects its shortages of those very things rather than a deliberate choice to retire the infantry.

  • Ben Sullivan

    Veteran journalist, 25 years ยท Science & business reporting ยท Founded ScienceBlog.com

    Ben Sullivan is a veteran journalist with 25 years of experience reporting on science and business across the U.S. and Europe. His work has appeared in premier outlets, including The Economist, The New York Times Magazine, the Los Angeles Times, and Prognosis, an English-language newspaper published in Prague. A digital media pioneer, Ben founded ScienceBlog.comย and led it for two decades. Under his leadership, the site was named one of the best science blogs "in the known universe" by Popular Science and was featured on Nature's year-end list of top science news blogs. Sullivan has consulted for the U.S. Department of State, served on the board of directors of the Los Angeles Press Club, was awarded a National Press Foundation fellowship to study health insurance, and taught writing at Loyola Marymount University's Asia Media International program. He lives in Los Angeles.

    MuckRack โ†— ยท LinkedIn โ†— ยท Editorial Policy & Correctionsโ†—

    https://orcid.org/0009-0007-1842-5997

Cite This Page

"Ukraine’s Drone War Has Compressed Combat to Machine Speed." ScholarPeer, 26 July 2026, scholarpeer.com/ukraines-drone-war-has-compressed-combat-to-machine-speed/.

Download RIS · Download BibTeX