A 2020 New America report cataloged thirty-eight states with armed drone programs, twenty-eight with programs in development, and eleven that have used drones in combat. In less than four years since that report was published, drones’ rapidly growing influence on battlefields from Nagorno-Karabakh to Ukraine to Gaza has almost certainly increased states’ interest in developing armed drone programs. As such, existing programs are likely to expand, new programs created, and new states pursuing and using drones. This explosion in drone interest comes as global warfare is entering a third age of drone warfare, defined by autonomy, saturation attacks, increased precision and range, and full-spectrum drone warfare across land, sea, and air.
Increasingly, states are integrating multiple drones into drone swarms, which communicate and collaborate to accomplish shared objectives. At least eleven states have announced drone swarm programs—from Armenia and China to South Korea and the United States. In May 2021, Israel used Elbit Systems’ Legion-X drone swarm in combat against Hamas, with the drones searching for targets and relaying information to the Israel Defense Forces.
Like any military system that proves its worth, drone swarm technology is likely to proliferate globally—perhaps even faster than some high-end exquisite weapons and platforms. But that proliferation is likely to be erratic and take place over years or even decades. Some states might assemble hundreds or thousands of drones into integrated, multidomain swarms, perhaps to defeat enemy submarines or defend against amphibious landings. Others might content themselves with simple, unarmed drone swarms for intelligence gathering and disaster response, and only after a fair amount of prodding. Although options to curb drone swarm proliferation are limited, they do exist.
Proliferation Dynamics of Drone Swarms
There are significant unknowns that will influence the shape and pace of swarm proliferation—how effective drone swarms prove in combat, in which missions, and the requirements to build or buy the most effective drone swarms. Theoretically, however, drone swarms can be applied to almost any mission. Numerous drones might search out, overwhelm, and destroy enemy air defenses, paving the way for manned aircraft. Swarms would be especially well suited for attacking transport and landing craft because they are often lightly defended, with swarms of cheap unmanned surface vehicles spreading across maritime flashpoints and hitting transport craft at the water level while unmanned aerial vehicles strike the military assets packed onboard the transport craft. Or perhaps they will take on the tactical role of mortars—small, aerial drone swarms providing longer-range effects with more maneuverability and native intelligence collection capabilities. The proliferation dynamics will look different depending on a swarm’s purpose, because purpose will determine the technology required.
Proliferation will also depend on drone swarms’ forms. Swarms consisting of small, homogenous unmanned aerial vehicles are likely to proliferate most quickly. A simple drone swarm capability is not necessarily all that difficult. Almost a decade ago, a team at the Naval Postgraduate School used commercial off-the-shelf technology to build a swarm of fifty unmanned aircraft systems (UAS). Building or buying fifty UAS is relatively easy, and although getting them to work together is harder, it is not necessarily as hard as some might expect—the Naval Postgraduate School swarm’s drone communicated with one another through Wi-Fi. The manufacturer needs to ensure the signals do not interfere with one another, build the command-and-control systems, and create any necessary user interfaces for human command. That all requires technical know-how, but basic knowledge is not necessarily hard to acquire. The DoD Strategic Capabilities Office’s Perdix drone swarm, for example, was designed by MIT engineering students.
Simple swarms are increasingly available commercially, though not yet on any significant scale. In February 2024, Geoscan, a Russian drone manufacturer, released a drone swarm starter kit providing ten quadcopters to practice with swarm intelligence algorithms for collective navigation and search. In 2022, Red Cat Holdings announced a “4-Ship” product, providing a prepackaged system of six UAS for automated surveying and area photography. The system enables simultaneous use of four UAS, with the other two UAS replacing the others as they run out of power. That said, the commercial market for drone swarms is probably limited. Environmental monitoring, safety inspections, and light shows might all benefit from drone swarming, but they do not necessarily need it. Federal Aviation Administration Part 107 regulations also require operators to obtain a waiver to fly multiple drones at once, so the drone swarm hobbyist market in the United States will probably be small. Of course, as more states pursue drone swarms, their availability on the market may increase. In April 2022, the Israeli company Elbit Systems’ UK branch was selected to provide the British Ministry of Defense with five autonomous swarms of six drones each.
However, not all drone swarms are as simple as these small, mainly quadcopter-based commercial examples. The Air Force is developing highly sophisticated Collaborative Combat Aircraft like the Kratos XQ-58 Valkyrie that escort and support F-22 and F-35 fighters in achieving air dominance and other roles. The CCAs are estimated to cost between $20.5 and $27.5 million each, incorporating cutting-edge advancement in artificial intelligence and autonomy. The Air Force reportedly aims for these drones to be capable of swarming. Likewise, large, multidomain drone swarms integrated into a specially designed mothership for transport will typically require much greater technical skill and resources to develop than a small swarm of quadcopters. Building and sustaining such a swarm will also require significant logistics, production, and maintenance capability.
The common constraints of drone proliferation will necessarily also constrain drone swarms. Stealth drones designed with low-observable features and radar-absorbent materials will be much slower to proliferate, and even more so if they are also swarming capable. Although stealth could significantly increase the costs of drones, making them less attritable, that may be desirable for certain missions. Likewise, drones that require more financial resources, specialized technical know-how, and infrastructure will proliferate slower, and adding swarming capability necessarily adds additional requirements. Just as a lone-wolf terrorist could build a crude fixed-wing drone in his garage, but not an MQ-9 Predator with a 66-foot wingspan, states with resource constraints might be able to field small swarms of inexpensive drones that communicate in rudimentary fashion, but comparatively few will manage to develop large and technologically advanced swarms with exquisite capabilities.
Managing Proliferation
Global proliferation of basic drone swarms should be assumed, but the proliferation of more complex drone swarms with advanced capabilities should not—at least not soon. And while proliferation cannot be completely controlled, the United States can take some actions to shape how quickly drone swarms proliferate. In cases where proliferation is all but certain—small, homogenous, aerial drone swarms, for example—the United States might want to encourage proliferation to partners and allies. If these swarms are expected to be a prominent feature on future battlefields—and based on observations from Ukraine and elsewhere, this is precisely what we should expect—then there is advantage in fostering close partnerships with allied states to not only develop such swarms, but also develop effective and responsible practices surrounding their integration into force structures, organizations, concepts, and doctrine.
Swarms that could meaningfully change regional or global balances of power, or that would provide more advantage to adversaries than to the United States, however, are an entirely different category. For example, concerns have been raised that antisubmarine drone swarms coupled with artificial intelligence and data analytics could pose significant risks to nuclear submarines, creating windows of opportunity for decapitation strikes. Compared to basic swarms, the proliferation of which should be assumed, these potentially destabilizing swarms require much more deliberate policy attention. The United States might attempt to limit proliferation through multilateral export controls on swarming-capable undersea drones, for instance. At a minimum, significant caution should be exercised in sharing relevant technology and know-how. And because certain drone swarms may actually represent a new category of weapons of mass destruction, the global community might develop new international norms concerning their proliferation.
Drone swarms are here. And while it is too early to precisely forecast the speed and trajectory with which the technologies will advance and swarms will spread throughout the world, the United States must prepare for their proliferation now.
Zachary Kallenborn is the lead researcher at Looking GlassUSA, a UAS and counter-UAS consulting and investment firm. He is also affiliated with the Center for Strategic and International Studies, the Schar School of Policy and Government, the National Institute for Deterrence Studies, and the National Consortium for the Study of Terrorism and Responses to Terrorism. His extensive research, writing, and analysis on drone warfare, weapons of mass destruction, terrorism, and existential risk has received global news coverage and influenced a broad range of state, federal, and global security policies and strategies.
This article is derived from “The Plague Beckons: On the Proliferation of Drone Swarms,” recently published in the Global Policy – ORF special volume, Future Warfare and Critical Technologies: Evolving Tactics and Strategies.
Thank you to Josh “JB” Bryan for edits on the article. All remaining errors and awkward turns of phrase are the author’s own.
Image credit: UK Ministry of Defence
This used to be just nightmare sci-fi stuff.
On the other hand, the biggest battleships in history were sunk by swarms (of piloted aircraft then), so ….