Why NATO's air defences cannot stop cheap drones—and what it would take to change that
The alliance spent decades building systems to defeat expensive missiles. Now it faces swarms of $50,000 drones that exploit every assumption baked into Cold War-era doctrine. The arithmetic of exhaustion favours the attacker.
🎧 Listen to this article
The Arithmetic of Exhaustion
A Patriot battery costs somewhere between $400 million and $1.1 billion. A single interceptor missile runs $3 million to $7 million. The drone it might destroy—a modified commercial quadcopter or an Iranian-designed Shahed—costs between $35,000 and $200,000. Fire enough interceptors at enough cheap drones, and even the richest alliance in history runs out of missiles before the attacker runs out of targets.
This is not a bug in NATO’s air defence architecture. It is the architecture itself, optimised for a threat environment that no longer exists.
For decades, Western air defence planners designed systems to defeat expensive, sophisticated platforms: cruise missiles, ballistic missiles, advanced fighter aircraft. These threats came in small numbers because they cost enormous sums to produce. The calculus was straightforward—spend heavily on exquisite interceptors because adversaries could only afford a handful of exquisite threats. The cost-exchange ratio, that cold metric comparing what attackers spend to what defenders must spend to stop them, favoured the defender. Or so the doctrine assumed.
Then came the drone.
When Physics Betrays Doctrine
The technical challenge begins with radar. NATO’s integrated air and missile defence systems rely on sensors designed to track objects with substantial radar cross-sections—fighter jets, cruise missiles, incoming ballistic warheads. A commercial drone presents a radar signature between 0.0001 and 0.01 square metres, roughly equivalent to a large bird. Advanced radars can detect such targets. Distinguishing them from actual birds, weather clutter, and ground reflections is another matter entirely.
The problem compounds at low altitude. Terrain masking, urban clutter, and atmospheric effects degrade radar performance precisely where small drones operate best. A Shahed-136 cruising at 200 metres above agricultural land exploits every limitation baked into systems designed to track missiles at 30,000 feet.
Speed creates its own paradox. Slow-moving drones—many commercial platforms fly at 50-80 kilometres per hour—confuse tracking algorithms optimised for fast-movers. The Doppler shift that helps radars distinguish missiles from background noise barely registers when the target moves slower than highway traffic. Some drones have learned to exploit this, flying patterns that mimic bird flocks or hovering at the edge of detection thresholds.
Then there is the engagement geometry. A Patriot battery positioned to defend against ballistic missiles cannot simultaneously cover the low-altitude approaches favoured by drone swarms. The radar looks up. The threat comes in low. Different systems serve different threat bands, but NATO’s layered defence concept assumes adequate coverage at every altitude. In practice, short-range air defence—the systems designed for low, slow targets—remains chronically underinvested across the alliance.
The Doctrine Gap
NATO’s February 2025 Integrated Air and Missile Defence Policy acknowledges the threat explicitly. It calls for protection against “small, low- and slow-flying UASs” alongside traditional missile threats. The document mandates “a mix of short-, medium- and long-range systems” providing “mutual and complementary support.”
The words are correct. The capabilities lag years behind.
For two decades, NATO members systematically divested short-range air defence assets. The Cold War ended, and with it the perceived need for SHORAD battalions protecting every brigade. Armies traded air defence units for deployable infantry. The threat from above would be handled by air forces achieving air superiority, or so the thinking went.
Ukraine has administered a brutal correction. Since September 2022, Russia has conducted at least 25 massive attacks on Ukrainian energy infrastructure, employing over 1,400 missiles and 500 strike drones. Ninety-eight percent of those drones were Iranian-designed Shaheds. Ukraine intercepts many—claims vary between 70% and 90% depending on the attack—but “many” is not “all,” and critical infrastructure requires near-perfect defence.
The pattern reveals a deeper doctrinal failure. NATO’s air defence architecture assumes a clear distinction between military and civilian airspace, between wartime and peacetime, between the defended zone and the undefended hinterland. Drones dissolve these categories. They operate in ambiguous legal space, launch from ambiguous locations, and strike targets that traditional military doctrine barely considers.
Who defends a power substation? The air force? The army? The interior ministry? The utility company itself? In most NATO states, the answer remains unclear. Institutional responsibility fragments across agencies designed for different missions, creating what analysts call “orphaned threats”—dangers that fall between bureaucratic jurisdictions.
The Production Arithmetic
Even if NATO solved the detection problem, even if doctrine caught up with reality, the numbers would still defeat the defender.
Consider the cost-exchange in its starkest form. A CSIS analysis found that Patriot interceptors cost $3 million to $7 million each, while typical drone targets cost $50,000 to $200,000. At the low end, defenders spend 15 times more than attackers. At the high end, 140 times more. No military budget survives that ratio indefinitely.
The disparity extends beyond unit costs to production capacity. Western defence contractors optimise for performance, not volume. They build small numbers of exquisite systems requiring specialised components, cleared facilities, and years-long production cycles. Drone manufacturers—whether state-backed factories in Iran or commercial operations in Shenzhen—optimise for scale. They use commercial off-the-shelf components available from global supply chains. They iterate rapidly. They produce in thousands.
China’s drone production capacity exceeds America’s by an order of magnitude. This is not speculation; it is industrial fact. The same factories that produce consumer quadcopters can pivot to military variants with minimal retooling. The same supply chains that deliver components for hobbyist drones serve military programmes. The distinction between civilian and military production—so central to Western defence industrial policy—barely exists in adversary states.
Directed energy weapons promise a solution: lasers that can destroy drones at pennies per shot rather than millions per missile. The technology works in testing. Deploying it creates new dependencies. Laser systems require substantial electrical power—power drawn from the very grid they are meant to protect. A drone attack that damages power infrastructure degrades the defences protecting power infrastructure. The loop closes on itself.
The Attacker’s Advantage
Adversaries have noticed the asymmetry and designed their tactics accordingly.
Russian doctrine now treats drones and missiles as complementary systems in combined attacks. Shaheds launch first, forcing defenders to expend interceptors and reveal air defence positions. Cruise missiles follow, targeting the gaps. The sequence exploits both the cost-exchange ratio and the reload time of defensive systems. A Patriot battery that has fired its interceptors at drones cannot engage the missiles that arrive minutes later.
Swarm tactics compound the challenge. Individual drones are easy to destroy; dozens arriving simultaneously from multiple vectors overwhelm any point defence. The coordination required is minimal—GPS waypoints and simple timing synchronisation suffice. The defender must track, prioritise, and engage every threat. The attacker needs only one to get through.
The innovation cycle favours offence. A drone design that fails can be modified within weeks using commercial components and 3D-printed airframes. A defensive system that fails requires years of procurement, testing, and deployment. The shanzhai innovation model—rapid iteration using available components without regard for intellectual property—enables attackers to adapt faster than defenders can respond.
Commercial supply chains enable this speed. The same GPS modules, flight controllers, and electric motors that power recreational drones serve military programmes. Export controls struggle to distinguish legitimate commercial use from weapons development. The components are too ubiquitous, the supply chains too distributed, the end-use too difficult to verify.
The Infrastructure Paradox
Critical infrastructure presents a defender’s nightmare: high value, fixed location, impossible to hide.
A power substation cannot be moved, dispersed, or made mobile. Its coordinates are public knowledge, often visible on satellite imagery. Defending it requires permanent presence—air defence assets that could otherwise be deployed flexibly must instead guard static positions. The attacker chooses when, where, and with what force to strike. The defender must be ready everywhere, always.
Ukraine’s experience illustrates the compounding effects. Successful strikes on power infrastructure reduce generating capacity, which reduces the electricity available for air defence radars and command systems. Damaged substations cannot be repaired quickly; specialised transformers have lead times measured in months. Each successful attack degrades not only the target but the broader system’s ability to resist subsequent attacks.
The value of infrastructure paradoxically prevents its protection. Active defence near power stations, refineries, or telecommunications hubs risks collateral damage from falling debris—interceptor fragments, drone wreckage, unexploded ordnance. Defenders face an impossible choice: engage threats and risk damaging what they protect, or hold fire and guarantee the attacker succeeds.
Some strategists argue for resilience over interception. If infrastructure cannot be perfectly defended, design it to fail gracefully. Distribute generation capacity. Build redundancy into grids. Harden critical nodes. Accept that some attacks will succeed and ensure the system survives anyway.
The approach has merit but faces its own constraints. Redundancy costs money. Distributed systems require more maintenance. Hardening takes years. And resilience, however robust, cannot prevent the psychological and political effects of successful attacks on visible targets.
The Institutional Mismatch
NATO’s structure compounds every technical and doctrinal challenge.
Air defence procurement remains a national responsibility. Each member state buys its own systems, sets its own priorities, maintains its own industrial relationships. Interoperability exists in principle; in practice, different nations operate different radars, different command systems, different interceptors. The alliance’s integrated air and missile defence coordinates these disparate capabilities but cannot compel standardisation.
Funding follows threat perception, and threat perception varies dramatically across the alliance. Baltic states that face Russian drones daily prioritise air defence. Western European members that have not experienced attacks prioritise other capabilities. The result is uneven coverage—robust in some sectors, threadbare in others.
Procurement timelines measured in decades cannot respond to threats evolving in months. A counter-drone system specified in 2020, contracted in 2022, and delivered in 2027 will face drones designed in 2026. The adversary’s OODA loop—observe, orient, decide, act—operates at commercial speed. NATO’s operates at bureaucratic speed.
The institutional architecture assumes clear boundaries: military threats handled by militaries, civilian protection handled by civilian agencies, peacetime distinguished from wartime. Drones respect none of these boundaries. They operate in gray zones—too small to trigger Article 5, too persistent to ignore, too ambiguous to assign clearly to any single institution.
What Changes the Calculus
Three developments could shift the balance, each with significant constraints.
First, directed energy weapons at scale. Laser and high-powered microwave systems eliminate the cost-exchange problem by reducing the marginal cost of engagement to near zero. The technology exists. Deploying it requires solving power generation, thermal management, and atmospheric interference challenges that have stymied programmes for decades. Progress is real but slow.
Second, autonomous counter-drone systems. AI-enabled sensors that can distinguish threats from clutter, and autonomous interceptors that can engage without human decision-making, could match the attacker’s speed of adaptation. The technology raises profound questions about human control over lethal systems—questions NATO has not resolved and shows little appetite to confront.
Third, offensive counter-drone operations. Rather than defending every potential target, strike the launch sites, production facilities, and supply chains that enable drone attacks. This requires intelligence, precision strike capability, and political will to conduct offensive operations that risk escalation. It also requires accepting that some attacks will succeed while offensive operations degrade the adversary’s capacity over time.
None of these solutions is quick. None is cheap. None eliminates the fundamental asymmetry between attackers who can fail repeatedly and defenders who must succeed every time.
The Uncomfortable Conclusion
NATO’s billion-dollar air defence systems cannot stop $500 drones because they were never designed to. They were designed for a different threat, procured through a different process, deployed according to a different doctrine. Adapting them to the current environment requires not incremental improvements but fundamental reconceptualisation of what air defence means, who provides it, and how it integrates with broader resilience strategies.
The alliance has begun this work. The 2025 IAMD policy acknowledges the threat. Counter-UAS programmes are accelerating. Short-range air defence is receiving renewed investment. But the gap between recognition and capability remains vast, and adversaries are not waiting for NATO to catch up.
The drone has not rendered traditional air defence obsolete. It has revealed the limits of systems optimised for threats that announce themselves through size, speed, and sophistication. The new threat is cheap, slow, small, and abundant. Defending against it requires not better versions of existing systems but different systems entirely—and different institutions to deploy them.
Until that transformation completes, the arithmetic will continue to favour the attacker. A $500 drone will remain capable of shutting down infrastructure that cost millions to build and billions to defend. The asymmetry is not a failure of technology. It is a failure of imagination—and imagination, unlike interceptor missiles, cannot be purchased at any price.
Frequently Asked Questions
Q: Why can’t NATO just use cheaper missiles to shoot down drones? A: NATO is developing lower-cost interceptors, but even “cheap” guided missiles cost tens of thousands of dollars—still unfavourable against $500 targets. The real solution requires non-kinetic systems like lasers or electronic warfare that reduce engagement costs to near zero, but these technologies face deployment challenges that will take years to resolve.
Q: How effective is Ukraine’s air defence against Russian drones? A: Ukraine claims intercept rates between 70% and 90% depending on the attack, but even 90% success means one in ten drones reaches its target. Against critical infrastructure that cannot tolerate any hits, this remains insufficient. Ukraine has also developed innovative low-cost solutions, including modified anti-aircraft guns and electronic warfare systems, that NATO is studying closely.
Q: Could commercial drone regulations help prevent attacks? A: Export controls and commercial regulations have limited effect because military drones increasingly use ubiquitous civilian components—GPS modules, flight controllers, electric motors—that cannot be practically restricted. The same supply chains serving hobbyists serve weapons programmes, making end-use verification nearly impossible.
Q: What would it cost to properly defend NATO’s critical infrastructure from drones? A: No comprehensive estimate exists, but the scale is daunting. Defending even a fraction of Europe’s power substations, refineries, and telecommunications nodes would require thousands of short-range air defence systems, tens of thousands of trained operators, and ongoing interceptor procurement that could exceed current defence budgets. Most analysts argue resilience—hardening infrastructure to survive attacks—offers better value than attempting perfect defence.
Sources & Further Reading
The analysis in this article draws on research and reporting from:
- NATO Integrated Air and Missile Defence Policy (2025) - NATO’s official doctrinal framework for addressing drone and missile threats
- CSIS: Cost and Value of Air and Missile Defense Intercepts - Detailed cost-exchange analysis of interceptor economics
- Cost-exchange ratio (Wikipedia) - Technical explanation of the military metric central to this analysis
- ACAPS/Dixi Group: Russia’s attacks on Ukraine’s energy sector - Comprehensive documentation of infrastructure attacks
- Advancements in Drone Detection Technology - Technical analysis of radar limitations against small drones
- Defense Security Monitor: Loitering Munitions - Industry analysis of drone and loitering munition development
- NATO Critical Infrastructure Security Research - Academic analysis of infrastructure protection challenges
- Electronic Component Supply Chain Analysis - Industry perspective on component availability and supply chain vulnerabilities
