Defense Tech: what are the biggest unsolved problems?

Last updated: 28 October 2026
Cover of NewMarketPitch's Defense Tech Reality Check report, 2026 edition Does Anduril take most of the defense tech money? Not most, but by far the biggest share. Anduril alone took 20% of the $45.0B raised by defense startups recently. With nine others, the top 10 take 56%. Chart from NewMarketPitch's Defense Tech report.How much does it cost to shoot down a drone? Often far more than the drone. A Stinger missile costs ~$480k, as much as ~14 Shaheds. Interceptor drones now stop Shaheds for $2,000 to $15,000, and lasers are tested at a few dollars a shot. Chart from NewMarketPitch's Defense Tech report.Beyond the US, where does defense tech money go? To Germany first. Outside the US, German startups took 48% of the $12B raised recently, mostly Helsing, Quantum Systems and Stark. That is 1.3 times the rest of Europe together. Chart from NewMarketPitch's Defense Tech report.When did money start pouring into defense tech? In 2025, and it sped up in 2026. Defense startups raised $5.0B in 2024, $11.5B in 2025 and $28.5B in 2026 to 14 November, already more than twice 2025. Chart from NewMarketPitch's Defense Tech report.How long would startup money fund US defense? About 11 days. Defense startups raised ~$30B in the last 12 months, while the Pentagon’s 2026 budget comes to ~$2.8B a day. Europe and Canada’s extra NATO money, ~$300B a year, would last ~4 months. Chart from NewMarketPitch's Defense Tech report.Do battle-tested Ukrainian startups get the money? Barely. Ukraine built ~4 million drones in 2025, yet its startups raised only ~$92M in disclosed rounds recently. That is ~1.5% of what German startups raised. Chart from NewMarketPitch's Defense Tech report.Are defense companies priced like big tech? Not the old ones. The primes are worth ~1.7 to 2.8 times their yearly sales, Europe’s war winners ~3 times. Anduril’s last round values it at 28 times its 2025 sales, Shield AI’s new one at ~96 times an estimate of its sales. Chart from NewMarketPitch's Defense Tech report.Who builds the most drones? Ukraine. Its industry now builds about 10 million drones a year, its ministry says, after ~4 million in 2025. Russia builds ~6,000 Shahed-type drones a month, and the US ~100,000 drones a year. Chart from NewMarketPitch's Defense Tech report.

Every funding round, every investor, and who is really ahead in defense tech, in one document. Have a look ↓

Summary

The biggest unsolved problems in defense tech are affordable protection against drones and missiles, producing weapons fast enough to replace losses, operating without reliable GPS or communications, and deploying trustworthy autonomous systems. Affordable defense and rapid replenishment are the two most urgent challenges.

Record military budgets have not translated into faster deployment. Global military spending reached $2.887 trillion in 2025, yet major U.S. weapons programs still take more than 12 years on average to deliver an initial capability. The gap between battlefield innovation and government procurement is widening.

The economics of warfare have shifted against defenders. Cheap drones can force militaries to deploy expensive layered defenses, with the Congressional Budget Office estimating $74 million to protect a representative military installation and at least $5 million annually to operate the system.

Missile defense has an equally serious replenishment problem. CSIS estimated that U.S. Patriot interceptor inventories had fallen below 1,000 by July 2026, with THAAD inventories around 250. Advanced defenses may work extremely well during an attack while consuming ammunition that takes years to replace.

Manufacturing capacity is improving, but unevenly. European annual 155mm artillery ammunition capacity increased roughly four to six times between 2020 and 2025. Complex missiles and warships remain much harder to scale, and announced factory capacity does not necessarily translate into delivered weapons.

Electronic warfare is changing what military autonomy needs to accomplish. Drones that depend on constant radio control or satellite navigation are vulnerable to interference, pushing manufacturers toward fiber-optic connections, onboard navigation, and autonomous operation. Each workaround creates fresh engineering and operational problems.

Military AI is progressing faster than the institutions responsible for testing it. The Pentagon is expanding AI applications while GAO has documented incomplete oversight responsibilities, staffing shortages, and reduced independent testing capacity. The hardest question is whether these systems can remain reliable when conditions are unfamiliar or deliberately deceptive.

Autonomous swarms and directed-energy weapons could eventually change the economics of combat, but demonstrations are running ahead of proven large-scale operational capability. Coordination, electronic warfare, weather, power requirements, and human supervision still limit what these technologies can achieve.

Defense startups are becoming serious competitors in autonomy, drones, and software. Their advantage is faster iteration, but scaling factories and supporting military equipment for decades exposes them to many of the same problems facing established contractors. Open interfaces and modular procurement may matter as much as the identity of the supplier.

The central lesson is that military capability must be judged over the length of a conflict, not during its first successful deployment. The strongest defense technologies will be those that remain affordable, repairable, adaptable, and available after communications fail, inventories shrink, and the enemy changes tactics.

Written by the New Market Pitch teamWe track new markets so founders and investors can move faster. About us

Why are defense technology problems getting worse despite record military spending?

Defense technology is struggling to keep up with modern warfare, even as governments spend more money on weapons than ever before.

According to the Stockholm International Peace Research Institute (SIPRI), global military spending reached $2.887 trillion in 2025, up 2.9% after adjusting for inflation. It was the eleventh consecutive annual increase. European military spending grew 14% to $864 billion, while Ukraine increased its military expenditure by 20% to $84.1 billion.

Yet the battlefield has changed faster than many military procurement systems. In Ukraine, inexpensive drones have become essential weapons against tanks, artillery, vehicles, and personnel. In January 2026, President Volodymyr Zelenskyy said drones accounted for more than 80% of enemy targets destroyed by Ukrainian forces. That figure comes from Ukraine's own reporting, but it reflects how dramatically the role of drones has expanded.

Meanwhile, conflicts involving Iran have demonstrated the enormous cost of defending against repeated missile attacks. The United States and its allies have used advanced interceptors whose inventories can take years to replenish.

The industry now faces several problems at once. Weapons are becoming cheaper and easier to deploy, but protecting against them remains expensive. Military hardware is becoming increasingly sophisticated, while factories struggle to produce enough equipment. And software is improving quickly, often much faster than governments can approve and deploy it.

Is the biggest problem in defense tech actually getting new weapons into service?

Getting defense technology into service remains one of the industry's biggest problems, particularly for major weapons programs that take more than a decade to deliver.

The U.S. Government Accountability Office (GAO) reported in July 2026 that the Department of Defense expected to spend more than $2.4 trillion on its most expensive weapons programs. The average time needed to deliver an initial capability had climbed to more than 12 years, compared with approximately 10 years in a 2019 assessment.

That is a worrying direction. Despite years of procurement reforms, the average delivery timeline has become longer.

The Pentagon's Replicator initiative illustrates a similar problem with newer technology. Launched in August 2023, the program aimed to deploy thousands of relatively inexpensive autonomous systems within two years. A Congressional Research Service update in January 2026 noted that a former defense official had reported only hundreds fielded by the original deadline. The complete final deployment figures were not publicly established.

Traditional procurement often begins by specifying detailed requirements long before the equipment enters service. Manufacturers then build around those requirements, while changes can trigger additional reviews, testing, and costs.

This approach becomes particularly difficult for software-driven weapons. Companies can update drone navigation software or AI models several times during the period required to approve a major hardware modification.

In its July 2026 annual assessment, GAO found that major U.S. defense programs continued missing important milestones despite repeated attempts to accelerate procurement. The Defense Department had increased its use of faster acquisition pathways intended to deliver capabilities within five years, but some programs entered those pathways with technologies that were still too immature.

Some flexibility is necessary because military systems must meet demanding safety, security, and reliability standards. But procurement becomes counterproductive when compliance procedures prevent militaries from adopting technologies that are already working elsewhere.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

What are the biggest unsolved problems in defense tech? Getting bought, building at scale and surviving jamming. Big US weapon programs take 12+ years to deliver, China refines 91% of the rare earths in motors and magnets, and late budgets slowed 36 of 74 programs. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

This is just a preview page. Don’t worry if a number looks old: the report gets refreshed, and you’ll always get the latest version.

Why is defending against cheap drones still so expensive?

Counter-drone defense remains a major unsolved problem because inexpensive drones can force militaries to spend millions on detection equipment, interceptors, and permanent defensive infrastructure.

In July 2026, the Congressional Budget Office estimated that equipping a representative U.S. military installation with comprehensive layered drone defenses would cost approximately $74 million, plus at least $5 million annually to operate. Protecting 100 similar installations would require around $7.4 billion upfront and $500 million per year.

Those estimates describe complete defensive systems rather than the cost of shooting down one drone. Still, they reveal the scale of the problem: even protecting fixed military facilities against small drones can require billions of dollars.

Militaries have several ways to respond. Traditional missiles can destroy aircraft but are expensive and limited in number. Electronic warfare can interfere with drone communications, although some autonomous and fiber-optic-controlled drones are harder to disrupt. Interceptor drones are potentially cheaper, but reliably catching fast or maneuvering targets remains challenging.

Companies such as Fortem Technologies, Anduril, and Epirus are pursuing different solutions, including autonomous interceptors and high-power microwave systems.

The Congressional Budget Office concluded that no single counter-drone technology provides complete protection. The most promising approach combines several defensive methods and chooses the cheapest effective response for each threat.

Counter-drone technology Main advantage Biggest remaining problem
Missile interceptors Effective against demanding aerial threats High cost and limited inventories
Interceptor drones Potentially cheaper per engagement Reliability against fast and numerous targets
Radio-frequency jamming Can disrupt drones without destroying them Limited effect on autonomous and fiber-optic drones
High-power microwaves Potential to affect multiple electronic targets Limited effective range and variable target vulnerability
Laser weapons Very low marginal energy cost per shot Weather, tracking and power requirements
Guns and programmable ammunition Established short-range defense Limited range and ammunition consumption
Passive defenses Reduce exposure without repeated interceptions Cannot prevent every attack
Layered defensive networks Combine complementary technologies High installation and maintenance costs

Are Patriot and THAAD missile defenses running out of ammunition?

Patriot and THAAD air defenses are facing serious inventory pressure, and recent conflicts have shown how quickly advanced interceptors can be consumed.

A July 2026 analysis by the Center for Strategic and International Studies (CSIS) estimated that U.S. Patriot interceptor inventories had fallen below 1,000, while THAAD inventories were around 250. These were analytical estimates rather than official stockpile disclosures, but they highlighted a major vulnerability after extensive missile-defense operations involving Iran.

Production has historically been much slower than wartime consumption. According to CSIS, U.S. procurement averaged approximately 270 PAC-3 MSE interceptors annually between fiscal years 2015 and 2024.

During the June 2025 Israel-Iran conflict, more than 150 THAAD interceptors were reportedly fired, according to subsequent CSIS analysis. In only twelve days, the number of interceptors expended was equivalent to a substantial share of the system's historical production.

The problem becomes particularly serious when several regions need the same defensive weapons. Interceptors committed to the Middle East cannot simultaneously protect forces in Europe or the Pacific.

Manufacturers can expand production, but specialized components, testing requirements, and long manufacturing timelines limit how quickly inventories recover. There is no quick fix when a weapon takes years to manufacture and can be consumed in minutes.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

How much of the “up to” money is really committed? Very little so far. Under Anduril’s $20B US Army framework, at least $144M has been ordered, 0.7%. Saronic is the exception, with about half of its Navy deal obligated. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

Just a preview. If a number here looks old, no worries: your report comes with the latest data.

Why can't defense manufacturers produce weapons fast enough?

Defense manufacturers still struggle to increase production quickly because missiles, ammunition, ships, and military electronics depend on specialized factories, workers, and components that cannot be expanded overnight.

Europe's artillery industry shows how much progress is possible. According to a 2026 assessment by the International Institute for Strategic Studies, NATO European countries increased their estimated annual 155-millimeter ammunition production capacity from 300,000-400,000 rounds in 2020 to approximately 1.6-1.8 million rounds in 2025.

That represents an increase of roughly four to six times in five years. Capacity could exceed 3.5 million rounds annually by 2030 if planned expansions succeed.

The improvement is substantial, although production capacity should not be confused with actual deliveries. New factories need reliable supplies of explosives, propellants, metals, and other materials. They also need skilled workers and customers willing to commit to long-term purchases.

Shipbuilding demonstrates an even harder version of the problem. In March 2025, GAO reported that the U.S. Navy had failed to increase its fleet size over approximately twenty years despite almost doubling its shipbuilding budget. Its Constellation-class frigate program was expected to deliver its first ship at least three years late after construction began before the design was complete.

These examples show different sides of the same difficulty. Artillery capacity can increase significantly when governments commit resources over several years. Complex military platforms are harder to accelerate because their designs, suppliers, and construction processes are deeply interconnected.

Can military drones survive when enemies jam their communications?

Military drones remain highly vulnerable to electronic warfare, although newer navigation methods and fiber-optic communications are making some systems harder to disrupt.

Russia and Ukraine have repeatedly interfered with each other's drone communications and satellite navigation. This has encouraged manufacturers and frontline operators to adapt their equipment, sometimes within weeks rather than years.

Fiber-optic drones are one response. Instead of transmitting control signals through radio frequencies, these drones remain connected to their operators through thin cables. Conventional radio jamming becomes much less effective, although the cables introduce physical limitations and handling problems.

Other systems increasingly rely on onboard navigation, image recognition, or automated flight functions. These approaches reduce dependence on continuous radio contact but require more processing power and dependable sensors.

The trade-offs are difficult. A remotely controlled drone benefits from human judgment but depends on communication. A more autonomous drone can continue operating after losing its connection, yet its behavior becomes harder to supervise when the environment changes unexpectedly.

Can newcomers end the missile shortage with cheap missiles? Not yet. Newcomers signed deals for thousands of cheap missiles a year, but the only firm order made public is for 50 Castelion missiles. The old makers deliver, 620 Patriot interceptors in 2025, and scale up slowly. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

This is a preview page. Spotted an outdated number? That’s normal here. The report you get is always the freshest version.

Why is the military still so dependent on GPS?

Military dependence on GPS remains a dangerous weakness because aircraft, drones, precision weapons, ships, and communications networks all rely on positioning or timing information that adversaries can disrupt.

An October 2026 GAO report confirmed that U.S. efforts to modernize military GPS continued to face major delays. After years of development problems, the Defense Department canceled its multibillion-dollar Next Generation Operational Control System in 2026 and shifted toward upgrading the existing system.

GAO also reported that more than two decades of work and billions of dollars had still not delivered the full intended military GPS modernization capability.

The consequences are already visible in Ukraine. Russian electronic warfare has degraded the effectiveness of some Western-supplied precision-guided weapons by interfering with satellite navigation signals.

Alternative technologies exist. Inertial navigation can track movement without constant satellite updates, while terrain-based navigation can compare sensor observations against known geographic features. Other systems combine multiple satellite networks or use backup timing sources.

Each alternative has limitations. Inertial systems accumulate errors over time, terrain-based methods depend on suitable environmental information, and alternative satellite signals can face similar interference.

Can military AI actually be trusted on the battlefield?

Military AI is already useful for analyzing information and assisting operators, but we still lack enough evidence to trust autonomous systems with unrestricted battlefield decisions.

The biggest concern is how AI behaves outside familiar testing conditions. Military environments contain camouflage, destroyed infrastructure, civilians, damaged sensors, misleading information, and deliberate attempts to confuse automated systems.

An image-recognition model might correctly identify a vehicle during testing but struggle when the same vehicle is partially hidden, modified, or surrounded by unfamiliar objects.

The U.S. military is expanding its use of artificial intelligence, yet its own oversight structures remain incomplete. In an October 2026 report, GAO found that the Defense Department had not fully resolved responsibilities for responsible AI implementation. As of February 2026, only three of ten authorized positions in its Responsible AI Division were filled.

That staffing problem does not establish that deployed AI systems are unsafe. It does show that the organization responsible for guiding responsible adoption has lacked some of the resources needed to do its work.

Testing is another concern. In June 2026, GAO found that staff reductions had weakened the Pentagon's independent operational testing oversight, leaving some personnel responsible for more programs than they could effectively manage.

Companies such as Palantir, Anduril, and Shield AI are developing different military AI capabilities. Their applications range from intelligence analysis to autonomous flight. The standard of reliability, and the supervision required, should rise sharply with the consequences of failure.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Are defense startups keeping their promises? Not all of them. Of 9 public promises, 1 was kept, 3 slipped or are disputed and 5 are still open. Factories arrive on time. Weapons that must pass tests or work at the front slip. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

Preview only. Some numbers may have changed since we made this page, but your report will be up to date.

Are autonomous drone swarms ready for real combat?

Autonomous drone swarms are progressing, but large groups of drones operating together without constant human supervision remain difficult to deploy reliably in combat.

The concept is attractive. Instead of controlling individual drones separately, an operator could assign a mission to several systems that coordinate their movements, share information, and adapt when individual aircraft are lost.

A coordinated group must avoid collisions, distinguish friendly and hostile equipment, manage limited communications, and continue operating when some drones lose navigation or sensor capabilities. Every additional drone increases the number of interactions the system may need to manage.

The Pentagon's Replicator initiative has supported the acquisition of uncrewed systems, including platforms from AeroVironment and Anduril. It has also involved software companies working on autonomy and command systems. However, fielding hundreds of autonomous platforms does not establish that they can function as one large, coordinated swarm.

There is also an economic question. Inexpensive drones become less attractive if effective coordination requires costly processors, specialized communications infrastructure, and extensive human monitoring.

Why can't military systems from different manufacturers work together?

Military interoperability is still a major problem because sensors, weapons, aircraft, and software platforms often use different technical standards and communication systems.

Consider a modern air-defense network. Radar equipment may detect an incoming aircraft, but the information still needs to reach the appropriate command system and interceptor quickly enough to respond. If those systems use incompatible data formats or cannot share classified information, the overall defense becomes less effective.

The difficulty grows when several countries participate in the same operation. NATO members operate equipment purchased from different manufacturers and developed under different national requirements.

The U.S. military's Combined Joint All-Domain Command and Control initiatives have tried to improve information sharing across military services and operating environments. Similar integration challenges exist in European defense programs.

Companies such as Palantir, Anduril, and Helsing are developing software intended to help military organizations connect sensors, information, and operational systems. But interoperability also depends on manufacturers agreeing to technical interfaces and governments resolving security restrictions.

Proprietary systems create another difficulty. A customer that builds its entire military software environment around one supplier can find it expensive to introduce competing equipment. And in defense, switching suppliers can become a strategic problem, not just an expensive IT project.

Can defense supply chains survive a major geopolitical crisis?

Defense supply chains remain exposed to major disruptions because many weapons depend on materials and electronic components sourced from a small number of countries or specialized manufacturers.

China's position in rare-earth processing is one important example. Rare-earth materials are used in permanent magnets, electronics, motors, and other military applications. Even countries with domestic mineral deposits may depend on processing capacity located abroad.

Semiconductor production creates similar concerns. Military systems use a wide range of chips, from advanced processors for autonomous systems to older components embedded in equipment designed decades ago.

Replacing a supplier can be difficult because military components often require specialized qualification and long-term availability. A commercially available substitute may not meet the necessary reliability, temperature, security, or performance requirements.

GAO has repeatedly identified weaknesses in the Defense Department's management of critical materials. Its 2024 review found shortcomings in how the department assessed and responded to supply-chain risks. Follow-up work continued into 2026.

The problem extends beyond raw materials. Explosives, rocket propulsion components, specialized machine tools, and industrial chemicals can all become bottlenecks when military production increases suddenly.

Complete domestic self-sufficiency would be extremely expensive, particularly for smaller countries. A more achievable approach combines supplier diversification, strategic inventories, alternative materials, and agreements with trusted international manufacturers.

Can defense startups really compete with Lockheed Martin and RTX?

Defense startups can already challenge established contractors in drones, autonomy, and software, but competing across entire military programs is much harder.

Anduril has become a prominent example of a newer defense company developing autonomous aircraft, sensors, command software, and manufacturing capabilities. Shield AI has focused on autonomous flight and uncrewed aircraft, while Helsing has expanded in European military AI and autonomous systems.

These companies are entering markets that were historically dominated by established defense suppliers. Their advantage often comes from shorter development cycles and products designed around software updates.

However, companies such as Lockheed Martin, RTX, Northrop Grumman, General Dynamics, and BAE Systems operate on a different scale. Their businesses include combat aircraft, missiles, naval systems, electronics, and long-term maintenance programs.

A company capable of producing an effective autonomous drone still has a long way to go before it can manufacture and support advanced combat aircraft or large integrated weapons systems.

The competitive picture is also changing because startups are investing more seriously in factories and large production facilities. That brings them closer to traditional defense manufacturing, along with many of the same capital and supply-chain challenges.

We expect the strongest competition where governments can purchase modular technologies independently. Traditional contractors remain difficult to displace when a contract requires decades of maintenance, complex certification, and enormous industrial capacity.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Which defense budget lines are growing the fastest? Munitions, counter-drone and drones, in the US. Each is asked to grow 4 to 5 times in the 2027 request. Europe’s biggest budgets grow 8% to 29% a year. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

This is just a preview page. Don’t worry if a number looks old: the report gets refreshed, and you’ll always get the latest version.

Can military communications survive cyberattacks and satellite outages?

Military communications are still vulnerable to cyberattacks and infrastructure failures, especially as more operations depend on connected software, satellites, and remote systems.

One of the clearest examples occurred in February 2022, when a cyberattack disrupted Viasat's KA-SAT satellite communications network during Russia's invasion of Ukraine. The attack affected users across several European countries, showing how disruption to shared infrastructure can spread beyond the immediate military target.

Modern military networks have several points of exposure. Satellites can be interfered with, ground terminals can be compromised, software can contain vulnerabilities, and connected civilian infrastructure can become an indirect target.

Encryption protects information against interception, but it does not prevent a network from becoming unavailable. A highly secure communication system is still a problem if military units cannot use it during an operation.

Redundant satellite services, alternative radio networks, isolated systems, and local data processing can help. Yet every additional connection and fallback method introduces costs and technical complexity.

Can laser weapons finally solve the drone problem?

Laser weapons could make some drone interceptions dramatically cheaper, but their performance still depends heavily on weather, power, tracking, and the type of target.

The appeal is straightforward. Conventional missile interceptors are expensive and must be manufactured, stored, transported, and replaced. A high-energy laser uses electricity to damage a target, potentially allowing many engagements without consuming a missile for each one.

That makes directed energy particularly interesting against relatively inexpensive drones.

However, laser systems need a clear and stable line of sight. Fog, rain, dust, and atmospheric conditions can interfere with the beam. The system must also track the target accurately and keep energy concentrated on it long enough to achieve the intended effect.

High-power microwave systems approach the problem differently. Companies such as Epirus are developing systems intended to disrupt electronic components, potentially affecting multiple drones during an engagement.

These technologies could improve defensive economics, but their effectiveness depends on the target's characteristics and the operating environment. Neither provides reliable protection against every type of aircraft or missile.

Why do military robots still need so much human supervision?

Military robots still need substantial human supervision because autonomous movement and decision-making become much harder in unpredictable terrain and contested environments.

Drones have achieved significant operational adoption, but ground and maritime robots face different problems.

A ground robot moving through damaged buildings, mud, rubble, or dense vegetation needs to recognize obstacles and respond to conditions that may change constantly. GPS interference and poor visibility can make navigation harder.

Maritime robots face weather, waves, navigation hazards, and long periods away from direct human control. Underwater vehicles have even greater communication constraints because radio signals travel poorly through seawater.

Companies such as Milrem Robotics, Shield AI, and Saronic are developing autonomous systems for different environments. Their progress is meaningful, but successful demonstrations do not establish that the equipment can complete long missions without intervention.

The important measure is how often operators need to step in. If a supposedly autonomous vehicle requires constant remote supervision, some of its expected advantages disappear.

Which defense technology problems are the hardest to solve?

The hardest defense technology problems today are those that combine technical limitations with production costs, institutional delays, and unpredictable battlefield conditions.

Some problems already have workable solutions that need to be manufactured more widely. Europe has demonstrated that artillery ammunition capacity can increase severalfold over a relatively short period when governments support sustained industrial investment.

Other problems are harder to solve by spending more money. Military AI still requires better testing under unfamiliar conditions. GPS-independent operations involve major changes to navigation systems and equipment already in service. Counter-drone defenses must adapt continuously because attackers can change their weapons and tactics.

Procurement is particularly frustrating because many of its problems have been documented for decades. GAO's latest assessments show that major weapons programs continue to experience longer delivery timelines despite repeated reform efforts.

We compared the problems according to their documented severity, the maturity of available solutions, and the difficulty of deploying those solutions at scale. These are qualitative judgments rather than numerical rankings of military risk.

Defense tech problem Evidence of the gap Main obstacle Current assessment
Counter-drone defense $74M estimated to protect one representative installation Cost and changing threats Critical
Missile-defense inventories CSIS estimated fewer than 1,000 U.S. Patriot interceptors in July 2026 Production and replenishment Critical
Military procurement More than 12 years to initial capability for major programs Slow acquisition processes Critical
Ammunition manufacturing European 155mm capacity rose roughly 4-6x since 2020 Industrial capacity and suppliers Improving
GPS-independent navigation More than 20 years of military GPS modernization difficulties Technical integration and reliability Critical
Military AI reliability Incomplete Pentagon oversight and staffing Testing and accountability Unresolved
Autonomous drone swarms Limited public proof of large-scale coordinated combat autonomy Coordination and electronic warfare Early-stage
Defense interoperability Continued investment in joint command systems Incompatible platforms and data Persistent
Naval manufacturing U.S. fleet stagnation despite nearly doubled shipbuilding budgets Design, workforce and infrastructure Critical
Directed-energy defense Effectiveness varies across targets and conditions Physics and system integration Promising but limited
Which way of stopping drones is winning orders? Interceptor drones. Ukraine built ~100,000 in 2025 at ~$2,000 to $8,000 each, and startups win the orders. Lasers have one big order, £316M in the UK, and microwaves are still prototypes. Chart from NewMarketPitch's Defense Tech report.
Defense Tech reportThis page and 70+ morePDF Get the report

Just a preview. If a number here looks old, no worries: your report comes with the latest data.

What are the biggest unsolved problems in defense tech today?

The biggest unsolved problems in defense tech are affordable protection against drones and missiles, producing weapons fast enough to replace losses, operating without reliable GPS or communications, and deploying trustworthy autonomous systems. We consider affordable defense and rapid replenishment the two most urgent challenges.

The same weakness appears across very different military technologies.

The United States has invested heavily in sophisticated air-defense systems, yet CSIS estimates show how quickly their inventories can shrink during sustained missile attacks. The Congressional Budget Office has calculated that protecting just 100 representative installations against small drones could require billions of dollars. Meanwhile, GAO continues to document acquisition programs taking more than twelve years to deliver an initial capability.

The manufacturing picture offers some encouragement. European artillery capacity has increased dramatically since 2020, showing that governments and manufacturers can respond to urgent demand. But that progress required years of investment, and expanding production of complicated missiles, aircraft, and ships remains much harder.

Autonomy, resilient navigation, and directed energy could improve the situation. They may reduce operating costs, make systems harder to disrupt, and allow militaries to deploy equipment more flexibly. However, those advantages still depend on whether the technologies can perform reliably during prolonged operations.

Our final judgment is that the defense industry's biggest challenge today is keeping military capability affordable and available during prolonged conflict. Military advantage increasingly depends on how quickly forces can adapt, replace losses, and continue operating after their first layer of technology or infrastructure fails.

If you want more recent data on this point, see the latest figures in our Defense Tech report ↓.

Our methodology

This analysis examines the biggest unresolved problems in defense technology using evidence available through October 2026. We divided the industry into distinct operational, technical, industrial, and institutional challenges, selecting problems supported by documented battlefield vulnerabilities, procurement delays, manufacturing constraints, integration failures, or persistent technical limitations. We focused on problems with measurable consequences rather than technologies attracting the most investment or publicity.

We distinguished fundamental technical obstacles from deployment problems. Some technologies still struggle to perform reliably in demanding conditions, while others have demonstrated workable capabilities but remain constrained by manufacturing, procurement, integration, or cost. For emerging technologies such as autonomous drone swarms, military AI, robotics, and directed energy, we distinguished successful demonstrations, early deployments, and sustained operational capability. We gave greater weight to evidence of performance under realistic conditions than to isolated demonstrations or manufacturer claims.

We assessed urgency by examining immediate operational exposure, the consequences of failure, replacement timelines, and the difficulty of restoring lost capability. We examined defense economics beyond individual weapon prices, including infrastructure, operating costs, ammunition consumption, replenishment, and scalability during prolonged conflict. For manufacturing, we treated announced production capacity, actual deliveries, and operational demand as separate indicators. This distinction is particularly important when comparing improvements in European artillery capacity with shortages of advanced interceptors and delays in naval construction.

Our final comparison is qualitative. Procurement delays, interceptor inventories, production capacity, and AI reliability measure different kinds of problems, so we did not combine them into an artificial numerical score. Instead, we compared what each indicator revealed about the severity, persistence, technological maturity, and scalability of the underlying capability gap. The assessments in the comparison table reflect these judgments, rather than a formal ranking of military risk.

We prioritized recent government assessments, official program records, independent defense research, and documented operational outcomes. Estimates from research organizations are identified as estimates, while battlefield statistics attributed to national authorities are treated as reported operational figures rather than independently verified global measures. We also considered limitations in publicly available information on military inventories, autonomous-system deployments, and classified capabilities when interpreting the evidence.

Key sources used for this analysis include: SIPRI's 2025 global military expenditure assessment, the Ukrainian presidency's January 2026 statement on drone use, GAO's July 2026 Weapon Systems Annual Assessment, GAO's review of defense acquisition reforms and persistent delays, the Congressional Research Service's January 2026 Replicator review, the Congressional Budget Office's July 2026 counter-drone defense cost analysis, CSIS's July 2026 assessment of Patriot and THAAD inventories, CSIS's analysis of missile-defense replenishment, IISS's 2026 research on European ammunition production capacity, IISS's assessment of munitions consumption and industrial output, GAO's March 2025 naval shipbuilding assessment, GAO's October 2026 military GPS modernization review, GAO's October 2026 report on responsible military AI, GAO's June 2026 review of operational testing oversight, NATO's Data Strategy for interoperability, NATO's 2026 interoperability exercise report, GAO's review of critical-material supply-chain risks, Viasat's original KA-SAT cyberattack incident report, and the UK National Cyber Security Centre's assessment of the same incident.