Military strength has traditionally been measured in factories, weapons and defense budgets. Increasingly, it also depends on something less visible: a nation's ability to turn scientific discoveries and commercial technologies into effective military capabilities.
Russia's war against Ukraine has exposed a paradox at the heart of modern military power. A country can mobilize enormous resources to manufacture weapons, replenish ammunition stocks and sustain a long war while struggling to develop the next generation of military technology.
Russia has shown that it can expand weapons production, adapt existing systems and build foreign components into its military equipment. Yet its difficulty obtaining advanced electronics, retaining scientific talent and developing its civilian economy raises questions about whether it can keep improving its technology over the long term.
The issue extends far beyond Russia. During much of the Cold War, the federal government was the principal funder of American research and development. Today, businesses fund most U.S. R&D, creating an enormous pool of technology that armed forces can buy, adapt and integrate into their operations.
Private enterprise has not replaced government as the engine of military innovation. The shift has instead created a new strategic challenge: how well can a nation connect scientific research, commercial ingenuity, industrial production and military requirements?
The countries best positioned for future conflicts may not be those that spend the most on defense or invent the most sophisticated individual weapons. They may be those that can consistently turn available knowledge and technology into reliable, affordable and useful military capabilities.
The Great Shift in Research and Development
The changing source of American research funding illustrates the transformation. According to the U.S. National Center for Science and Engineering Statistics (NCSES), the federal government funded about 66.8 percent of U.S. research and development in 1964, and businesses about 30.8 percent. The two lines crossed around 1980. By 2024, the estimated shares were about 18.7 percent federal and 74.8 percent business.1
| Year | Federal funding share | Business funding share |
|---|---|---|
| 1964 | 66.8% | 30.8% |
| 1980 | 47.4% | 48.9% |
| 2024 (estimated) | 18.7% | 74.8% |
Shares of total U.S. domestic R&D funding, not military research alone. Remaining shares come from universities, nonprofits and state governments.
The shift is unmistakable, but the numbers need two qualifications.
First, they describe who pays for research, not who performs it. During the Cold War, private companies already performed most American R&D, including work the government paid for. In 1964, businesses performed about 68.3 percent of U.S. R&D.2 Government agencies, universities, laboratories and contractors were parts of one interconnected system.
Second, the federal share fell mainly because business research grew enormously, not because Washington withdrew. Federal R&D spending has continued to rise. Between 2023 and 2024 alone, adjusted for inflation, business funding grew by about $17 billion and federal funding by about $7 billion.3
What has changed is that businesses are now the dominant source of R&D money as well as its largest performers. The commercial sector invests enormous sums in computing, artificial intelligence, semiconductors, biotechnology and communications, all fields with military applications. Modern armed forces can therefore draw on a technological base that extends far beyond traditional defense contractors and government laboratories.
From Military Invention to Commercial Innovation
During the early Cold War, the United States invested heavily in technologies that commercial markets could not yet support on their own. Competition with the Soviet Union drove work on advanced electronics, communications, navigation, aerospace engineering and computing.
Many of these investments eventually reached civilian life. The internet grew from government-supported networking research. GPS began as a military navigation system before becoming indispensable to transportation, logistics and smartphones. Semiconductors benefited from both private invention and government demand, including the needs of the space program.
This pattern became known as spin-off: technology developed for military or government purposes spreading into the civilian economy.
Today the flow increasingly runs the other way. Companies develop technologies for huge civilian markets, and militaries adapt them. This reverse process is called spin-on.
Consider the difference between building a computer network for a military research program and drawing on a mature commercial cloud industry. The second option gives defense organizations years of investment, engineering refinement, cybersecurity development and operating experience without having to recreate any of it. Similar opportunities exist in satellite communications, software, artificial intelligence, robotics and unmanned aircraft.
Commercial markets also create a cycle of improvement. Companies sell to many customers, collect feedback, release new versions and reinvest the revenue. Competition rewards experimentation, and the scale of civilian demand lowers costs.
Militaries can benefit from that cycle, but commercial technology does not automatically become a military advantage. It must be adapted, tested, integrated and supported under conditions very different from those it was designed for.
The Commercial Battlefield
The war in Ukraine has offered some of the clearest demonstrations of civilian technology on the battlefield. Three areas stand out: drones, satellite communications and software.
1. Drones: Cheap Technology, Rapid Adaptation
Commercial components have made small drones widely available to armed forces. Cameras, batteries, electric motors, flight controllers and radios built for hobbyist and civilian markets can be assembled into inexpensive reconnaissance and attack systems.
Ukraine has built extensive networks linking military units, engineers, commercial suppliers, volunteer groups and domestic manufacturers. Through these networks, frontline users report problems and work directly with developers to modify equipment.
The resulting feedback loop can be remarkably fast. A drone runs into electronic interference, a control system proves unreliable or a mission changes. Engineers develop a fix, units test it and the results shape the next version. The process has made unmanned systems a defining feature of the war. Small drones can scout, locate targets, direct artillery and strike vulnerable equipment at a fraction of the cost of traditional weapons.4
Their effectiveness still has limits. Drones can be jammed, grounded by weather and shot down before reaching their targets. Their value depends on trained operators, reliable communications, suitable payloads and the ability to replace losses. And they work alongside artillery, infantry and armor rather than replacing them.
Russia has adapted too. Its fiber-optic-controlled drones send control signals through a thin cable instead of a radio link, which makes them far harder to jam, though the cable limits their range and maneuverability. Battlefield innovation is not an exclusive Ukrainian or Western advantage.
The broader lesson: access to cheap technology matters, but so does the capacity to adapt it, manufacture it in quantity and answer an opponent's countermeasures.
2. Satellite Communications: A Strategic Dependency
When ground communications are destroyed or disrupted, military units and government agencies can struggle to stay connected. Low-Earth-orbit satellite services can restore connectivity across damaged or contested areas, and Ukraine's use of Starlink has shown the value of commercially built communications infrastructure in a major war.
The same example reveals the weakness of the commercial model. In 2022, SpaceX chief executive Elon Musk declined a Ukrainian request to activate Starlink coverage near Sevastopol for an attack on Russia's Black Sea Fleet, as he later acknowledged publicly. (Early reports that he switched off coverage already in place were corrected by his biographer.)5 Whatever one thinks of the decision, a private executive's judgment shaped a military operation.
A commercial service can be highly capable without being under a nation's sovereign control. Commercial providers can deliver technology faster and sometimes more cheaply than traditional defense programs, but armed forces must also plan for redundancy, alternative suppliers, cybersecurity and the possibility of losing access in a crisis.
Commercial satellite imagery offers a related opportunity. Private companies can supply additional observations of troop movements, infrastructure and battlefield conditions, supplementing national reconnaissance satellites, though weather, revisit rates, processing capacity and access restrictions limit their value.
3. Software and Artificial Intelligence: Turning Information into Decisions
Modern warfare generates enormous quantities of information. Satellite images, drone video, radar tracks, intercepted communications and reports from troops must be interpreted fast enough to inform decisions.
Commercial software and artificial intelligence can help organize that information, spot patterns and speed the work of analysts. U.S. programs such as Project Maven show defense organizations adapting machine learning to intelligence work. Commercial cloud and cybersecurity services can also keep government operations running under attack. Ukraine's early move of key government data and services onto resilient infrastructure showed that civilian technology can contribute to national survival as well as battlefield effectiveness.
But buying a technology is not the same as proving it works. A contract shows that an organization has acquired a capability; it does not show that the capability improves targeting, speeds decisions or produces better outcomes. Artificial intelligence can make mistakes, produce misleading output or fail when conditions differ from its training. Military users must be able to check its results, protect sensitive data and keep human judgment in charge.
Russia's Dilemma: Producing Weapons Is Not the Same as Innovating
Writing for the Jamestown Foundation, Paul Goble argues that Russia now has two economies: a war-driven military-industrial sector and a struggling civilian one. Borrowing émigré analyst Sergey Rogov's term "bipolar," he contends the divide points to disaster ahead, because in most other countries innovation now comes from the civilian sector rather than the state.6 That raises a central question: can a country sustain a powerful defense industry while its civilian economy and independent science stagnate?
Russia has redirected enormous resources toward the war. Its defense industry has expanded production of ammunition and other equipment, adapted existing weapons and developed modifications to meet battlefield demands. These achievements are real. Russia retains considerable industrial capacity and should not be dismissed as incapable of technological adaptation.
But expanding production and creating new generations of technology are different challenges. A factory can raise output by adding shifts, securing components and building more copies of an established design. Developing new electronics, sensors, propulsion systems or autonomous platforms takes years of research, specialized talent, advanced manufacturing equipment and repeated testing.
Russia faces constraints in each of these areas. Chatham House research by Mathieu Boulègue describes how sanctions and wartime demands have compounded long-standing weaknesses in Russian defense innovation, while noting that Russia retains enough capacity to keep fighting.7 Restrictions on international scientific cooperation and the emigration of researchers and engineers weaken parts of the innovation base. Centralized procurement makes it hard for independent developers to bring new ideas into service.
Russia's reliance on imported electronics shows the limits of self-sufficiency. When RUSI researchers examined 27 Russian military systems recovered in Ukraine, they found hundreds of foreign-made components, most of them from Western suppliers.8 A country can assemble sophisticated weapons while still depending on others for essential parts.
Such dependencies are not unique to Russia; every advanced economy relies on foreign suppliers. The strategic question is whether a country can keep access to critical inputs, develop substitutes and continue improving its systems when supplies are cut.
Here the evidence supports a more measured conclusion than Jamestown's. Russia's economic troubles should not be confused with imminent military-industrial collapse. A weakening civilian economy can narrow future technological options while the state keeps prioritizing weapons production and sustains the war. The long-term danger is not that Russia will suddenly lose the ability to build weapons. It is that its ability to develop, absorb and scale the next generation of technology may fall behind its competitors'.
Ukraine's Lesson: Innovation Depends on Feedback
Ukraine offers a contrast, though not a simple story of private enterprise defeating centralized government.
One of Ukraine's notable strengths has been the close contact between frontline users and the people building their equipment and software. Engineers, volunteer groups, startups, commercial suppliers and government programs have helped connect battlefield needs with technical solutions.
The artillery and mapping system Kropyva is an example. It grew out of Ukraine's volunteer defense-technology effort, associated with the organization Army SOS, and runs specialized targeting software on ordinary tablets and laptops.9 Its development shows how digital tools can be shaped around practical military problems, from different ammunition types to different weapons. Ukraine's government-backed BRAVE1 initiative, launched in 2023, formalizes the same idea by connecting defense-technology developers with military users.
The advantage of such arrangements is not decentralization for its own sake. It is the shorter distance between spotting a problem, testing a solution and putting an improvement into service.
Traditional military procurement emphasizes formal requirements, extensive testing and large contracts. Those procedures are essential for complex weapons but poorly suited to technologies that change quickly and are needed urgently. A system that allows controlled experimentation, rapid feedback and the scaling of successful designs responds faster to changing conditions.
Still, Ukraine's innovation has depended on much more than its own technology community. Foreign military aid, imported components, industrial production, artillery ammunition, training and state coordination have all been essential. Ingenuity cannot substitute for ammunition, logistics, maintenance or production at scale. Innovation becomes military power only when it is connected to the institutions and resources needed to put it to work.
The United States: Technological Abundance, Institutional Challenges
The United States has an enormous commercial technology sector, world-class research universities, advanced manufacturers and a defense industry capable of building highly specialized systems. Yet having advanced companies does not mean the military will adopt their technology quickly.
The Government Accountability Office's 2026 annual assessment of major weapons programs found that the average time to deliver a new capability had risen to more than 12 years.10 A commercial company can release several generations of a product in that time. The gap is widest in software, autonomous systems and other technologies that evolve faster than traditional weapons programs.
The Pentagon knows this. Its November 2025 Acquisition Transformation Strategy made commercial products and services the default choice and expanded flexible contracting methods designed to bring in nontraditional suppliers.11 The Replicator initiative, launched in 2023 to field large numbers of inexpensive autonomous systems, has since been folded into a new Defense Autonomous Warfare Group,12 and the Defense Innovation Unit continues its work of bringing commercial firms into defense contracting. The repeated reorganizations are themselves evidence of how hard the integration problem is.
Announcing reforms is not the same as delivering faster, and the GAO report adds an important caution. Between 2018 and 2025, 18 of 40 programs entered the Pentagon's rapid-acquisition pathway with immature technologies, and that slowed delivery rather than speeding it up. Moving fast works best when the technology is ready.
Nor should the Pentagon simply buy whatever Silicon Valley offers. Commercial products may not survive electronic attack, extreme environments or the loss of civilian infrastructure. Military-specific engineering, secure systems, specialized sensors and nuclear technologies will continue to require substantial government involvement.
The American challenge is therefore one of integration: preserving the strengths of public research and established defense contractors while making far better use of commercial innovation. That requires procurement systems that reward successful experimentation, test technology under realistic conditions and move proven solutions into service, along with long-term support for maintenance, training, cybersecurity and manufacturing.
China's Alternative: A Hybrid Innovation System
China complicates any argument that commercial innovation must come from a predominantly private-sector system.
China combines large commercial technology companies, universities, research institutes, state-owned defense enterprises and extensive industrial policy. Its military modernization draws on both civilian technology and state-directed research and production.
A 2025 study by Georgetown University's Center for Security and Emerging Technology examined thousands of artificial-intelligence contract awards by China's People's Liberation Army. In a sample of repeat contract winners, close to three-quarters were nontraditional vendors with no reported state-ownership ties, though state-owned enterprises and defense-affiliated research institutes still led the list of most frequent winners.13 Civilian firms can contribute to military technology even where the state plays a major coordinating role.
Procurement records have limits, however. They count contracts, not spending, and they cannot show how well systems perform or whether a particular government policy caused the results. The study's authors make this point themselves.
China's experience shows that the relationship between civilian enterprise and military power can take different institutional forms. The United States and China organize their innovation systems differently, but both must solve the same underlying problem: turning technological potential into operational capability.
Why Private Enterprise Cannot Do Everything
The growing importance of commercial technology should not obscure the continuing role of government.
Companies invest where they expect a commercial return. That drives rapid progress in computing, communications, robotics and other fields with large markets. But many important scientific advances have uncertain applications, long timelines or benefits no single company can capture.
That is why government still funds a far larger share of basic research than of R&D overall. Although Washington paid for only about 18.7 percent of total U.S. R&D in 2024, it funded an estimated 39.9 percent of basic research.14 Public procurement can also create demand for capabilities that have no civilian market yet.
The history of semiconductors illustrates the interaction. Private firms developed key components, government and space-program demand drove their use in demanding applications, and commercial markets then expanded the industry far beyond its original customers.
Military equipment also faces demands commercial products are not designed for. It must work in severe environments, withstand deliberate interference, integrate with other systems and remain available in wartime. And depending on a few commercial providers carries risk: companies can change priorities, restrict access or be disrupted, and concentrated supply chains become strategic weaknesses when a critical part cannot be replaced quickly.
The strongest model, then, is not one in which government steps back and leaves innovation to business. It is one in which public research, commercial experimentation, defense-specific engineering and military adoption reinforce one another.
The Real Measure of Military Innovation
Competition over military technology is often described as a race to invent the next breakthrough. But invention is only the first stage.
A country must develop or obtain a useful technology, manufacture or buy it in sufficient quantity, integrate it into military organizations, train people to use it and maintain it in the field. Then it must keep adapting as opponents develop countermeasures. Failure at any stage can waste the original innovation.
A sophisticated weapon produced in small numbers may matter less than a simpler one built at scale. A cheap drone may prove useless if its signals are easily jammed or its operators poorly trained. An advanced AI system may contribute little if commanders cannot fit it into how they make decisions.
Conversely, a modest improvement can have large consequences when it is affordable, reliable and quickly adopted. Russia's glide bombs are the clearest example. By fitting old Soviet-era FAB bombs with UMPK wing-and-guidance kits, Russian forces gained a cheap stand-off strike weapon, used at scale from late 2023, without needing new aircraft or an entirely new family of precision munitions.
Military innovation should therefore not be measured by scientific novelty alone. Adaptation, manufacturing capacity and operational usefulness matter as much. The same applies to commercial technology: a country does not gain a decisive edge merely because its companies make the world's most advanced software, chips or drones. It gains an edge when those technologies become effective military capabilities and its institutions keep improving them as the battlefield changes.
Conclusion: The New Arsenal Is an Ecosystem
The transformation of military innovation is not a simple contest between government laboratories and private companies, and it does not prove that one national economic model will inevitably defeat another. Public and private institutions have always worked together on important technologies. What has changed is the scale of the commercial technology base available to modern armed forces.
That base creates an advantage only when institutions can use it. Governments must sustain foundational research, manufacturers must produce reliable equipment, procurement systems must adopt useful technology and military units must learn to employ it. Industrial scale, logistics, manpower, doctrine, training and strategic choices remain fundamental to the outcome of any war.
The new arsenal is not simply a collection of weapons or a network of factories. It is the entire ecosystem that turns scientific knowledge and commercial ingenuity into capabilities that can be produced, deployed, sustained and improved.
In future conflicts, the decisive advantage may belong not to the nation that invents the most impressive technology, nor to the one that builds the most weapons, but to the one that most effectively connects invention, production and military learning.
Notes
- National Center for Science and Engineering Statistics, National Patterns of R&D Resources: 2023–24 Data Update, NSF 26-313 (February 2026), Table 6, "U.S. R&D expenditures, by source of funds and performer: 1953–2024." Table 6 (PDF). 2024 figures are estimates. ↩
- NCSES, National Patterns of R&D Resources: 2023–24 Data Update, NSF 26-313, Table 2, "U.S. R&D expenditures, by performer and source of funds: 1953–2024." Table 2 (PDF). ↩
- NCSES, "U.S. R&D Totaled $937 Billion in 2023; Estimate for 2024 Indicates Further Increase to $993 Billion," NSF 26-314 (February 27, 2026), figure 2 (constant 2017 dollars). ncses.nsf.gov/pubs/nsf26314. ↩
- For a field-based survey of how both armies adapted their tactics, including their use of drones, see Jack Watling and Nick Reynolds, Tactical Developments During the Third Year of the Russo–Ukrainian War (Royal United Services Institute, February 2025). RUSI. ↩
- "Elon Musk says he denied Ukraine satellite request to avoid complicity in 'major act of war' vs. Russia," CBS News, September 2023, CBS News; on the biographer's correction, "Walter Isaacson Corrects Reporting on Elon Musk Cutting Ukrainian Military's Access to Starlink," Mediaite, September 2023, Mediaite. ↩
- Paul Goble, "Russia's Bipolar Economy Opening Door to Disaster Ahead," Eurasia Daily Monitor, Jamestown Foundation, October 13, 2026. Jamestown Foundation. ↩
- Mathieu Boulègue, Russia's Struggle to Modernize Its Military Industry, research paper (Chatham House, July 21, 2025). Chatham House (PDF). ↩
- James Byrne et al., Silicon Lifeline: Western Electronics at the Heart of Russia's War Machine (Royal United Services Institute, August 2022). RUSI. ↩
- "Інновації для армії. Система «Кропива» та карти Армії SOS" [Innovations for the army: the Kropyva system and Army SOS maps], Ukrainska Pravda Life, August 11, 2015 (in Ukrainian). Ukrainska Pravda. ↩
- U.S. Government Accountability Office, Weapon Systems Annual Assessment: Requiring Mature Technologies Could Enable Shift to Rapid Delivery, GAO-26-108457 (July 2, 2026). GAO. ↩
- "Defense Department's Acquisition Transformation Strategy," USNI News, November 11, 2025 (full text of the strategy). USNI News. ↩
- "The Pentagon's $54 Billion Bet on Autonomous Warfare," Defense One, May 2026. Defense One. ↩
- Cole McFaul, Sam Bresnick and Daniel Chou, Pulling Back the Curtain on China's Military-Civil Fusion: How the PLA Mobilizes Civilian AI for Strategic Advantage (Center for Security and Emerging Technology, Georgetown University, September 2025). CSET. ↩
- NCSES, National Patterns of R&D Resources: 2023–24 Data Update, NSF 26-313, Table 7, "U.S. basic research expenditures, by source of funds and performer: 1953–2024." Table 7 (PDF). ↩
No comments:
Post a Comment