We are entering an epoch in which robots will not only fight but also account for much of the arithmetic of war. Procurement budgets, production lines, attrition rates, sustainment contracts, spare part logistics, and algorithmic decision trees will conspire to produce what I call a robot economy of war: a set of market and institutional behaviours in which economic incentives shape the character, tempo, and ethics of armed conflict.
Two obvious facts frame the argument. First, state spending on military capacity is large and rising, creating demand for new classes of systems. That demand is both the carrot and the currency for firms that supply autonomy, sensors, and munitions. Second, militaries are explicitly pursuing “attritable” and mass-producible unmanned systems as part of force design. Together these dynamics make it plausible that future conflicts will be fought by swarms and fleets bought, produced, and sustained according to market rules as much as strategic logic.
Consider the procurement logic. When a state treats a platform as attritable it signals a tolerance for loss. Attritable systems are intentionally lower-cost, designed to be expendable at the tactical level and therefore priced to enable mass employment. The U.S. Department of Defense has explicitly translated this into initiatives to field large numbers of inexpensive autonomous capabilities rather than a few exquisite platforms. That shift alters the buyer-seller relationship. The customer is less interested in longevity per platform and more interested in throughput: speed of production, modularity of payloads, and ease of field updates. Suppliers respond by optimizing unit cost, supply-chain resilience, and manufacturing velocity rather than marginal increases in per-unit performance.
The arithmetic of attrition is where the robot economy bites. Historic cost comparisons between uncrewed and manned platforms show that some unmanned systems offer much lower acquisition and operating costs per flight hour. That makes them attractive when commanders need persistent presence or high sortie rates. But operational economics are not only about cost per hour. They are about sustainment, software updates, parts, and replacement rates. A lower per-unit purchase price can be overwhelmed by high attrition rates or by the long tail of software maintenance and cybersecurity costs. Public budgets and private investors who buy into the dream of cheap mass-production can quickly be surprised by maintenance bills and accelerated obsolescence.
The battlefield in Ukraine has already offered a laboratory of constrained evidence. Private entrepreneurs, small workshops, and state programs combined to scale drone production and improvisation. That industrial agility matters. When a theater produces many thousands of low-cost attack drones, the defender faces a new calculus: intercepting a low-cost attacking drone with an expensive interceptor or airframe is an unsustainable exchange rate. The result is not merely tactical; it is economic coercion. The attacker uses cheap marginal costs to impose sustained expenditure on the defender, forcing changes in doctrine, procurement, and political will. Reporting from conflict zones and investigative pieces in open media have documented how this cost asymmetry can shape campaign choices and force structure decisions.
From an economic perspective there are several emergent market structures to watch.
1) Vertical integration and military-civil fusion. Dual use components from civilian markets will continue to dominate margins for many robotic systems. Sensors, processors, batteries, and radio modules are procured from commercial supply chains. This reduces marginal costs and speeds iteration, but it also exposes militaries to global supply fragility and raises questions about which firms sit at the chokepoints of war-time production.
2) Platform-as-product and software-as-service models. Expect more contracts that buy capability rather than hardware. Sellers will invoice for payload suites, persistent cloud models for perception stacks, and secure update services. The vendor who controls the update pipeline accrues market power because they can push fixes, new tactics, or even behavioral policies into fielded fleets. That creates governance issues: who asserts legal authority when a vendor patch changes lethality parameters in theater?
3) Markets for attrition. If attritable fleets are affordable, buyers will purchase by the thousand. That creates commodity markets where price per unit matters more than capability per unit. Variants will commodify into tiers: disposable loitering munitions at one end and expensive long-endurance autonomous vehicles at the other. Market mechanisms such as learning curves, economies of scale, and supply-chain localization will rapidly compress prices for low-end robotic ordnance.
4) Countermeasures as recurring revenue. Defense against low-cost robotic threats will be a perpetual and profitable market. Governments will buy layered systems that include kinetic interceptors, electronic warfare, jamming, and directed energy. Each solution carries its own lifecycle costs and capital intensity. The net effect is an arms market in which attacking robots are inexpensive marginally and defenders must either invest in cheaper counters or accept escalating expenditure.
These structures create perverse incentives. A supplier whose business model thrives on high churn of attritable systems may prefer sustained conflict over rapid resolution. A state that gains strategic leverage by producing cheap attacking robots may be less willing to de-escalate because its marginal cost advantage yields immediate operational leverage. We must recognise that markets do not have moral preferences. They align incentives in predictable ways. This is a normative problem dressed as an economic prediction.
Sustainment economies deserve special attention. Hardware is one component. Software updates, secure data-links, engines, and spare parts are another. Early comparisons between manned and uncrewed systems show attractive operating cost profiles for some UAS types, but those studies also warned about hidden life-cycle costs: software upgrades, cybersecurity hardening, sensor refreshes, and spare-part logistics. A fleet whose attrition rate is unknown will generate unpredictable sustainment spending that can eclipse acquisition savings. The buyer who plans around a fixed unit price without factoring in a reasonable replacement and maintenance curve risks severe budgetary surprise.
There are social and legal externalities that cannot be priced easily. Autonomy changes attribution and accountability. When a lethal action is the result of a chain combining a fielded algorithm, a vendor patch, and a tactical commander, who is responsible for unlawful harm? Market arrangements that allocate risk to private vendors can produce moral hazard if states outsource political accountability along with technical capability. The robot economy will thus strain legal norms of responsibility unless regulatory frameworks define red lines and liability structures. Academic and policy communities must push for procurement clauses that preserve investigatory access, incident reporting, and auditability of decision logs.
Finally, what should policymakers and scholars do now?
1) Treat procurement as a public policy instrument not merely a transaction. Contracts must require transparency for software updates, supply-chain mapping, and an auditable chain of decision-making for lethal effects.
2) Build economic models of attrition. Governments need plausible scenarios that link rates of loss to sustainment budgets and to substitution strategies. Realistic models will help avoid surprise when a cheap swarm forces expensive defensive responses.
3) Invest in cheaper, layered defenses and domestic surge manufacturing. If the market will produce cheap attackers, the defender must be able to produce or field cheap counters quickly. That is not simply a technological problem. It is an industrial policy problem.
4) Reframe arms control debates. Treat the economics of production as part of arms control. Limiting proliferation of specific components, enforcing export controls on critical sensors or guidance systems, and international transparency about production capacity are all pragmatic levers.
5) Preserve human judgment where it matters. War will always be a moral activity. Economic incentives will push toward automation where they reduce costs. Where the stakes are highest, policy must ensure the human element remains central in decisions that bear on life and political consequence.
The robot economy of war is not a deterministic prophecy. It is a set of emergent tendencies rooted in current spending patterns, acquisition doctrines that prize attritable mass, and the real-world experience of low-cost unmanned systems on modern battlefields. If we surrender the economic levers of procurement and sustainment to market imperatives alone we will wake up to campaigns shaped less by strategy and more by unit price lists and supply-chain resilience. The choice is ours. We can shape the market with law, with industrial policy, and with ethics, or we can be shaped by it.