The Software Ate the Battlefield 

How NATO's $40 Billion Drone Push Signals a New Logic of War

Muhammet Talha Kaan 

 

 

NATO’s $40 billion Drone Edge is three admissions at once: that cheap, smart machines have ended the West’s monopoly on precision, that power inside the alliance now flows to whoever adapts fastest rather than spends most, and that the decision to kill is quietly passing from the operator to the algorithm.

In early 2026, the world’s most expensive military did something telling: it copied its enemy’s cheapest weapon. The American LUCAS drone — roughly $35,000 apiece, reverse-engineered from Iran’s Shahed — went into combat in the Middle East within days of its debut. Months later, on 7 July, NATO’s leaders gathered in Ankara and launched Drone Edge: more than $40 billion for counter-drone capabilities over five years, and five times as many drone operators by the end of 2027. The alliance’s official text states, flatly, that drones have become “a decisive factor on the battlefield.” It is a procurement programme presented as routine. It is better read as a confession — three of them, folded into one press release. And what changed the battlefield was not the drone. It was the software inside it.

THE END OF A MONOPOLY

For most of history, military power meant numbers. Since the Cold War it meant precision — hitting exactly what you aimed at, which only rich states could afford. Cheap drones, cheap sensors and, above all, cheap guidance software have now fused the two. Scholars Michael Horowitz and Lauren Kahn call the result “precise mass”: accuracy in bulk, available to anyone. Eric Schmidt, who has made repeated trips to Ukraine over the last four years, puts it more sharply. The fighting, he argues, “is no longer a war of platforms. It’s a war of systems” — the decisive unit is not the drone but the architecture that lets a force see, decide and strike faster than its enemy. The hardware is almost incidental; the edge lives in code. And code does not respect a defence budget.

The proof came from three theatres. In Ukraine, small attack drones now cause roughly 60–70 per cent of frontline casualties on both sides, and output climbed from about 800,000 in 2023 toward four million in 2025. In June 2025, Operation Spiderweb smuggled 117 drones costing $600–1,000 each deep into Russia and destroyed strategic bombers worth billions — at least ten of them, by satellite count. In the Middle East, Iran fired Shahed-type drones at American bases, Israel and the Gulf and exposed a brutal piece of accounting: killing a $20,000–50,000 drone with a Patriot missile costing over $4 million hands the attacker a ratio of eighty or two hundred to one. Hence LUCAS: if you cannot afford to shoot cheap drones down, you learn to fire them back. Then it came home — nineteen Russian drones crossed into Poland in September 2025, and unidentified ones shut Copenhagen’s airport for hours. The humiliation was not that NATO could not stop them. It was that it could not afford to. The West’s monopoly on precision was finished.

THE NEW HIERARCHY

Finished monopolies redistribute power — the second confession. If the edge lives in software rather than steel, the states that rise are the fastest to reorganise around code, not the richest; and that capacity is not distributed the way GDP is. Hence the summit’s unfamiliar cast. Turkey — the host, and no accident that the alliance’s centre of gravity now tilts toward its southern and eastern flank — fielded Baykar, which by its own government’s account exported $2.2 billion in 2025 and remained the world’s largest armed-drone exporter for a third year.^[3] Ukraine, not even a member, has become the alliance’s indispensable supplier of the one thing no factory produces — combat-tested know-how — and allies now treat it as a security provider, not a supplicant. Tiny Latvia co-leads a twenty-nation drone coalition with Britain.

This is not a shift away from America, whose technological weight is unmatched and which is racing the same way, aiming to buy 200,000 FPV drones of its own by 2027. It is a shift away from the slow — and Europe’s legacy giants are slow. Mario Draghi’s 2024 report found that 78 per cent of some €75 billion in EU procurement went to suppliers outside the Union, while the continent fields a dozen types of main battle tank to America’s one.^[5] The mindset is slower still: in March 2026 the chief executive of Germany’s Rheinmetall belittled Ukraine’s drone innovations — the authentic voice of an industry that prizes costly, exquisite platforms and mistrusts the cheap and the fast. Drone Edge is the first alliance programme to admit that its hierarchy now tracks adaptation, not wealth.

THE THIRD CONFESSION

The slow are punished; the fast automate. That is the third confession, and it is the one Drone Edge funds without daring to name. Look first at how desperately both sides are trying to keep a human in control. A drone’s radio link is its weakest point — jam it and the weapon drops — so Russia and Ukraine now fly drones on spools of hair-thin fibre-optic cable, physically tethered, impossible to jam. The counter is almost medieval: front-line roads are roofed in nylon nets and barbed wire to tangle the drones, and Ukraine plans thousands of miles of it. This is a war of algorithms, answered with chicken wire.

But tethering and netting are patches on the same fragile thing: the human on the end of the link. Remove the human, and the machine improves — autonomous navigation, CSIS researcher Kateryna Bondar found, lifts a drone’s success rate from around 10–20 per cent to around 70–80, by deleting the two weakest components, the radio link and the stressed operator. So the programme’s stated centrepiece — five times as many operators by 2027 — runs against its own logic. A strike-drone team needs three to six specialists; Ukraine trained over 5,000 last year and only 437 reached the top grade; and no hand on any joystick can meet a saturation swarm. The operator is not merely expensive. The operator is the bottleneck — and in a system built for speed, bottlenecks get engineered out.

Strategists have a name for where this ends. Chinese military thinkers anticipate a “battlefield singularity” — the point at which combat outpaces human cognition and decisions pass to machines — and warn Beijing may be readier to cross it. The competitive logic is merciless. As Robert Work, the former US deputy defence secretary, put it in Schmidt’s telling: “In the future, if the first person through the door is a human and not a robot, we are very stupid people.” NATO’s side is candid too: its innovation chief, Jean-Charles Ellermann-Kingombe, calls this “a tech race” and notes that Russia now iterates a design in as little as two to twelve weeks — against the twenty-four months NATO’s own Rapid Adoption Action Plan allows itself — before adding, of autonomy: “The way you can remove human beings from harm’s way, why wouldn’t you?”

THE SCEPTICS’ HOUR

The revolution thesis deserves its cross-examination, and serious scholars supply it. If Schmidt is precise mass’s most confident champion, its most formidable doubters are academics: Antonio Calcara and colleagues argue that drones have not revolutionised war at all, because air defence adapts and the old contest between hiding and finding endures. The numbers give them ammunition — by some assessments 60–80 per cent of small attack drones never reach their targets — and even CNAS, a home of the precise-mass argument, judges the change in Ukraine evolutionary rather than revolutionary. Cheap drones have burned bombers but won no one a strategic victory; the Donbas may not translate to the Pacific; and the $40 billion may prove less new money than it sounds, or may flow to the familiar giants rather than the proven newcomers.

All fair. But note what the sceptics concede. As Andriy Zagorodnyuk, Ukraine’s former defence minister, put it in a paper for the Carnegie Endowment, no one claims a single weapon has made all others redundant; what changes in a revolution is “the balance of capabilities” and the organisation built around it. That is exactly the claim here. The tank did not abolish infantry. It abolished the primacy of those who were slow to buy tanks.

So judge Drone Edge by its own tests: whether the adoption clock is met, whether the first contracts reward the proven or the merely established, and whether the cost of an interception ever falls below the cost of the thing intercepted.

In Ankara, the alliance formalised what Ukraine, Iran, and the drone incursions over Poland had already demonstrated: precision has been democratised, and power now belongs to those who can produce it cheaply, reorganise quickly, and automate first. NATO has acknowledged the first two imperatives; the third it is funding almost in silence. It is investing in a new military edge—but whether that edge ultimately belongs to human operators or to increasingly autonomous systems is the question Ankara left unanswered. The answer, however, is already being assembled, one procurement standard at a time. The last human in the loop will not be removed by decree. It will be priced out.