“In 2023 Ukraine produced around 800,000 drones. In 2024, two million. In 2025, up to five million. These are jaw-dropping numbers for somebody who’s been looking at defense technology.”
— Gregory C. Allen, CSIS, 2025
1.1 The Three Accelerants
For centuries, power belonged to those who could afford it. Empires built fleets. Nations built air forces. Only the wealthy could wage war. That equation is gone.
Today:
- $500 drone can destroy a $2 million tank
- $50 million program can inflict billions in damage
This is not an imbalance. It is a structural inversion.
The attacker has embraced disposability, the defender is still built for endurance. And in a world of mass production, disposable wins.
Three technological forces are converging to alter the frequency, duration, and cost of armed conflict. Each alone would be significant. Together, they constitute the most profound shift in the character of warfare since nuclear weapons.
The first accelerant is radical cost reduction. The primary weapon of modern warfare—the first-person-view drone—costs $300 to $2,000 to produce. Its targets—tanks, armored vehicles, radar installations—cost millions. This inversion of the traditional cost curve means that the financial barrier to effective military action has collapsed. A nation spending $50 million on drones can inflict damage previously requiring a $5 billion air force. Consider the arithmetic: Ukraine’s entire annual drone production budget—roughly $2–3 billion for 4 million units—is less than the cost of 40 F-35 fighter jets.
The second accelerant is artificial intelligence. AI compresses the time between detecting a target and destroying it from hours to seconds. AI-guided drones navigate, select targets, and strike autonomously, without communication links vulnerable to electronic warfare. In Ukraine, AI-enabled terminal guidance allowed drones to continue tracking targets even after Russian jamming severed all radio contact. By 2025, full-mission autonomy was under active development: drones using AI throughout their entire flight for navigation, target search, priority selection, and strike execution.
The third accelerant is industrial accessibility. Modern drone warfare does not require a superpower’s industrial base. It requires 3D printers, commercial electronics, open-source software, and small workshops—all available globally. Ukraine built an industry producing 4 million drones per year from a standing start in under three years, with over 300 defense-tech startups. Iran produces an estimated 10,000 Shahed loitering munitions per month. The barriers to entry have vanished.
1.2 The Collapse of Cost
When waging war becomes dramatically cheaper, the political calculus restraining conflict changes. Throughout the 20th century, the immense expense of deploying armored divisions, fighter wings, and naval task forces served as a natural brake. Leaders hesitated because the financial and political costs of failure were catastrophic. That brake is weakening.
The Houthis—an armed group in one of the world’s poorest countries—mounted a maritime campaign disrupting $1 trillion in global shipping with weapons costing less than many nations spend on a single fighter jet. The total cost of Houthi weapons used in the Red Sea campaign through 2024 has been estimated at under $100 million. The cost of the multinational defense against those weapons exceeded $1 billion. If a non-state actor can achieve strategic effects at this cost ratio, the incentive structure for armed conflict worldwide shifts toward greater frequency.
This does not mean every dispute becomes a war. Nuclear deterrence, economic interdependence, and alliance structures still restrain great-power conflict. But below the nuclear threshold—in the gray zone where drones operate—the friction preventing escalation is diminishing. Expect more proxy wars, cross-border drone campaigns, maritime disruptions, and internal conflicts fueled by cheap precision weapons. The cost of initiating conflict has fallen below the cost of many nations’ annual road maintenance budgets.
1.3 The Compression of Time
In the past, war moved at the speed of humans : Detection → Decision → Action
Minutes. Hours. Sometimes days. And today ? …. Seconds.
AI does not hesitate. It does not get tired. It does not misjudge under stress. And most importantly: It does not wait.
This creates a dangerous shift:
Whoever strikes first gains disproportionate advantage
Decision-makers have less time to think
Mistakes become instantaneous and irreversible
War is no longer just violent, It is accelerated beyond human rhythm.
AI-driven warfare is inherently faster than human-driven warfare. When targeting cycles compress to seconds, swarms saturate defenses in minutes, and critical infrastructure can be destroyed in opening hours, the decisive phase shortens dramatically. Israel’s systematic degradation of Iran’s air defenses over months meant that the 12-day June 2025 war’s outcome was shaped more by pre-conflict positioning than combat itself.
This creates first-strike pressure. If the side that attacks first with autonomous weapons gains decisive advantage—destroying air defenses, command networks, and retaliatory capability before the defender responds—the incentive to strike first in a crisis becomes overwhelming. The compression of decision timelines makes diplomacy harder and crisis instability greater.
However, shorter decisive phases do not necessarily mean shorter conflicts. Ukraine demonstrates that when both sides adopt drone warfare, the result can be protracted attrition—not because weapons are slow but because both sides regenerate capability fast enough to prevent knockout blows. Russia’s rate of advance averaged just 15–70 meters per day during its most prominent 2024–2025 offensives—slower than the Battle of the Somme in World War I. The conflict has produced nearly 1.2 million Russian casualties through December 2025, more than any major power in any war since World War II, yet Russia has gained less than 1.5 percent of Ukrainian territory since January 2024.
1.4 The Democratization of Destruction
Once, war required factories, states, and decades of preparation. Now it requires:
A workshop
A supply chain
And knowledge freely available online
Ukraine built millions of drones in under three years. Iran produces thousands monthly. And Non-state actors now operate at a strategic scale. The monopoly on violence has eroded.
A New Reality
When war becomes:
Cheap
Fast
Accessible
It does not just become more efficient.
It becomes more likely. Conflicts that once seemed too expensive. now fall below the threshold of hesitation.
The economics of warfare have inverted completely. The following figures illustrate the scale of the inversion:
Attack
Attack Cost
Defense Cost
Ratio (Attack:Defense)
Iran April 2024 barrage (320 weapons)
$35 million
$550 million
1:16
Single FPV drone vs. T-72 tank
$500–$2,000
$1.5–3 million
1:750–6,000
Shahed-136 vs. Patriot PAC-3
$20–50K
$4 million+
1:80–200
Houthi drone vs. SM-2 interceptor
$20–40K
$2.1 million
1:50–100
Houthi drone vs. SM-6 interceptor
$20–40K
$3.9 million
1:100–200
LUCAS drone (U.S.) vs. Iranian target
$35,000
N/A (offensive)
Cost parity with Shahed
Total Houthi campaign (est.)
<$100 million
$1+ billion (defense) + $1T disruption
1:10,000+
The LUCAS drone—America’s reverse-engineered Shahed—costs $35,000: the price of a mid-range sedan. Firestorm’s 3D-printed drones are manufactured in under nine hours. Ukrainian FPV drones cost less than a smartphone. When offense is this cheap and defense this expensive, the traditional model of deterrence through superior firepower erodes. The United States consumed roughly a quarter of its THAAD interceptor stockpile—representing years of production—in less than two weeks of combat with Iran. Only 37 THAAD interceptors were planned for FY2026 procurement; the war’s expenditure rate exceeded even peak historical production by 50 percent.
KEY INSIGHT: A $50 million drone program can now impose billions in damage and disruption. The cost of effective military capability has fallen by orders of magnitude, lowering the threshold for conflict worldwide. Iran’s six-hour April 2024 attack cost $35M; the defense cost $550M.
Metric
Traditional Era
Drone/AI Era
Change Factor
Cost per effective strike
$500K–$2M (cruise missile)
$500–$50K (FPV/Shahed/LUCAS)
100–4,000x cheaper
Design-to-deployment time
5–15 years
2–12 weeks
50–400x faster
Industrial base required
Major defense prime
3D printers + COTS electronics
Democratized
Minimum credible force budget
$10B+ annually
$100M–$500M annually
20–100x lower
Decision cycle (detect-to-destroy)
Minutes to hours
Seconds (AI-guided)
100–1,000x faster
Monthly production at scale
Hundreds (missiles)
200,000+ (drones, Ukraine)
1,000x+ higher
Interceptor:drone cost ratio
N/A
40:1 to 200:1
Defender bankrupted