Drone warfare refers to the use of unmanned aerial vehicles (UAVs) to conduct surveillance, reconnaissance, and lethal strike missions, often remotely piloted from thousands of miles away. When I think about drone warfare I think about the movie Eye in the Sky (2015), arguably the best fictional treatment of drone ethics. Helen Mirren and Alan Rickman star in a tense real-time drama about a strike decision with civilian collateral risk. It was excellent on the political and moral chain of command. Then the scariest one to me was Angel Has Fallen (2019), with Gerard Butler and Morgan Freeman where a drone swarm assassination attempt is a key opening sequence of terror and makes you realize how very hard it would be to defend against a drone swarm attack.
The concept of unmanned aircraft predates modern warfare significantly. In WWI, the U.S. developed the Kettering Bug (1918), a pilotless aerial torpedo that never saw combat. WWII saw further experimentation, including Germany’s V-1 “buzz bomb”, essentially a cruise missile, and U.S. attempts to convert B-17 bombers into radio-controlled weapons. During the Cold War, the U.S. flew reconnaissance drones over China and North Vietnam after losing manned U-2 spy planes, establishing the surveillance role that would define early drone doctrine. The real turning point came with the Predator drone, developed by General Atomics and first deployed by the CIA and Air Force in the Balkans in the mid-1990s for reconnaissance. It was unarmed, so basically a flying camera. After the 1998 embassy bombings in Africa, the CIA began exploring arming the Predator to target Osama bin Laden in there search for him. By 2000, test flights demonstrated the Predator could carry Hellfire missiles. The infrastructure for lethal drone strikes was in place before 9/11, but the attacks of September 11, 2001 transformed the drone from surveillance tool to primary strike weapon. The first armed Predator strike killed Mohammed Atef, al-Qaeda’s military chief, in Afghanistan. Then in 2002 the CIA conducted the first targeted killing outside a declared war zone, striking a vehicle in Yemen. Under Presidents Bush and especially Obama, the drone program expanded massively with strikes in Pakistan, Yemen, Somalia, Libya, and Syria.
Drone warfare introduced profound new questions. Arguments in favor of drones say they reduce risk to pilots, enable precision targeting, are cheaper than manned missions, and provide a persistent surveillance capability. The criticisms and concerns are over civilian casualties with “signature strikes” (targeting behavioral patterns rather than identified individuals) having killed unknown numbers of civilians in places like Ukraine. It is also a legal gray zones when it comes to strikes in non-war zones that challenge sovereignty and international law. The psychological impact of populations under constant drone surveillance in Pakistan and Yemen reported widespread anxiety and trauma and just look at how that felt recently here in California when it was suggested that Iran might be targeting us for drone attacks. Drones suffer from an accountability problem since they create a moral “distance” and lower the threshold for using lethal force. And drones are not like nuclear weapons… they have proliferated with over 100 countries now possessing military drones and non-state actors like ISIS and Houthi forces have weaponized commercial drones with apparent ease. The war in Ukraine has marked another evolution with cheap FPV (first-person view) drones costing a few hundred dollars and used as single-use kamikaze weapons alongside sophisticated systems, democratizing drone warfare to a degree never seen before. It has reshaped how armies think about every aspect of ground combat. Drone warfare has fundamentally altered the calculus of military force, lowering the cost and political risk of killing while raising deep questions about sovereignty, accountability, and the future of armed conflict.
And now the reality of swarm drones, as predicted in Angel Has Fallen, may be upon us. It may be cool when you see it at the Opening Ceremony at the Olympics, but not so much in attack formation. Swarm drone technology represents one of the most significant emerging developments in both military and civilian applications with the coordination of large numbers of small drones acting collectively, like a school of fish or a colony of ants. A drone swarm is a group of UAVs that operate collaboratively, sharing information and coordinating behavior like a flock of starlings. They operate either through centralized command, distributed AI, or fully autonomous decision-making (how much scarier can swarm drones get than autonomous killing?). The key distinction from simply flying many drones is emergent collective behavior: the swarm adapts as a system, not just as individual units. I’m sure that what they are not telling us that the swarm learns from every encounter and gets smarter and smarter about how to kill its prey autonomously…
As you can imagine, the U.S. Military with its DARPA OFFSET program has been developed swarm tactics for urban combat with hundreds of small drones working with ground forces. The Perdix micro-drone, developed by MIT students and tested by the Pentagon in 2016, demonstrated 103 drones launched from F/A-18s operating as a self-healing, adaptive swarm and became a landmark demonstration. The Replicator Initiative (announced 2023) aims to field thousands of attributable autonomous systems across all domains, directly responding to China’s mass production capacity in drones. China has demonstrated the largest public swarms with a 2017 test flying 1,000 fixed-wing drones simultaneously. CETC (China Electronics Technology Group) has shown swarms launched from helicopters and ground vehicles
Meanwhile, Russia has experimented with swarm attacks in Ukraine and they and the Ukrainians are getting pretty proficient about the technology. Russia has used Shahed-136 Iranian loitering munitions in swarm-like waves against Ukrainian infrastructure while Ukraine has struck Russian targets, including inside Russian territory, using waves of cheap FPV and fixed-wing drones that overwhelmed air defenses. Both sides have shown that even unsophisticated coordination of many cheap drones stresses expensive air defense systems enormously. Turkey’s Bayraktar ecosystem is moving toward coordinated multi-drone operations and Israel has been developing networked drone systems for years. Even Houthi forces in Yemen have used drone and missile swarms against Saudi infrastructure and Red Sea shipping, and you can bet there are a few swarms over the Strait of Hormuz as I write this.
Defending against swarms is genuinely hard. The economic asymmetry of swarm drones is a core strategic military problem when $500 drones can force the expenditure of a $1–3 million interceptor missiles… and the drones can outperform the missiles. Kinetic interceptors (missiles) are too expensive at scale. Directed energy (lasers) may be the most promising solution with the U.S. and Israel now deploying early systems, but range and power remain limiting. Electronic warfare / jamming can be effective, but swarms are being designed to operate without GPS or external signals so screw that solution. Sometimes net guns, trained eagles, and other kinetic, low-tech solutions are easier for limited scenarios. The real answer, though, seems to be AI-driven counter-swarms, fighting swarms with swarms, but there we are back at the AI conundrum. The autonomous weapons and related legal concerns just multiply. Fully autonomous defense swarms that select and engage targets without human decision-making raise profound questions, but they may be the only way to combat fully autonomous attack drone swarms. The international humanitarian law requiring distinction between combatants and civilians may be too much to expect for an AI driven judgment system. The U.S. currently requires a “meaningful human control” standard, but the definition of “meaningful” is contested and under pressure from operational realities. Do you wonder where Pete Hegseth ranks on the issue?
The technology is advancing faster than doctrine or law. Within the next decade, analysts expect swarms of thousands of coordinated autonomous drones to be operationally fielded, integration with ground robots and naval vessels for true multi-domain swarm operations, significant democratization with the barrier to entry continuing to fall, meaning more state and non-state actors will have meaningful swarm capability. AI will certainly enable real-time tactical adaptation that no human operator could match in similar speed. Swarm drones may ultimately represent as significant a shift in warfare as gunpowder, the machine gun or the tank, changing not just tactics but the fundamental structure of how conflicts are fought. That goodness man has learned that war and hostility don’t pay, right?