A drone swarm is a collection of cooperating autonomous drones under a common command or coordination algorithm. Unlike the mere parallel use of many individual drones, swarm systems exchange position, sensor, and target data among themselves and decide in a decentralized manner about the distribution of roles, attack vectors, and evasive maneuvers.
The major swarm programs of 2025/2026
The US Replicator program (announced 2023). Goal: field readiness of thousands of autonomous systems by August 2025. The deadline was missed; handed over to a Defense Autonomous Warfare Group in October 2025. In January 2026 an Orchestrator Prize Challenge with a prize of 100 million USD for swarm command software was announced (DoD press release).
Chinese swarm demonstration, January 2026. A demonstration by the People’s Liberation Army together with a national defense research institute showed how a single soldier steered over 200 fixed-wing drones by means of an adaptive, anti-jamming coordination algorithm. It is the largest public cooperative demonstration in the swarm domain to date.
Research swarm programs. Analogous programs in the EU (collaborative research projects within the framework of the EDF) and in Israel are documented. The combination of low per-unit costs, open-source software, and military demand portends an increasingly widespread capability.
The question of autonomy
Drone swarms are not necessarily fully autonomous weapon systems. They can operate at any of the four degrees of autonomy. The decisive question is: Who decides on the lethal application?
The practical tendency moves toward human-on-the-loop or human-in-command (a commander gives the swarm mandate, the algorithm decides on individual engagements). This is structurally the situation in which the responsibility gap appears especially sharply: not just one autonomous decision per mission, but hundreds in concert.
Personal-ontological assessment
The drone swarm intensifies the problems of the single system. Whereas with the autonomous single drone the question of attribution can still be focused on a concrete kill, in the swarm it systematically blurs:
- Who is responsible for a single killing if the swarm algorithm coordinated it?
- Who answers for collateral damage that arises from the emergent behavior of the collective?
- Is the commander, who issued only the mandate (“secure this sector”), responsible for every individual engagement decision?
Personal ontology draws the consequence: swarms with lethal effect must be built structurally such that personal ultimate responsibility remains locatable. Where this is architecturally not possible, deployment is impermissible. This is the military-technical concretization of the demand for meaningful human control.
Non-military swarms
Swarms are not exclusively military. Civilian applications range from light shows through agricultural sensing to search-and-rescue operations. These applications are unproblematic in personal-ontological terms, as long as they serve the dignity of the person and do not instrumentalize it. Surveillance swarms in police work, by contrast (facial recognition, movement profiling), can quickly tip over into instrumentalization — here the limits of the EU AI Act (Art. 5) and of fundamental-rights proportionality take effect.
Ontological classification
Superordinate concepts: entity (a collective compound system, not an individual agent)
Composed of: combat drones or civilian drones
Typical degrees of autonomy: human-in-command or human-on-the-loop at the swarm level, often fully autonomous at the individual level
Chapter assignment: Chapter 5: Oblivion of the Person (German)
See also
- Combat Drone
- Lethal Autonomous Weapon System
- Combat Robot
- Responsibility Gap
- Degree of Autonomy
- EU AI Act
- UN GGE LAWS
- War
- Robert Sparrow
Sources: Sparrow, Robert (2007): “Killer Robots.” Journal of Applied Philosophy 24(1), pp. 62—77.
Further sources:
- Scharre, Paul (2018): Army of None: Autonomous Weapons and the Future of War. W. W. Norton.
- United States Department of Defense (2023 ff.): Replicator Initiative — Press Releases and Congressional Testimony.
- United States Department of Defense (2026): Orchestrator Prize Challenge Announcement, January 2026.
- Kallenborn, Zachary / Bleek, Philipp C. (2018): “Swarming destruction: drone swarms and chemical, biological, radiological, and nuclear weapons”. Nonproliferation Review 25(5—6), pp. 523—543.
- Horowitz, Michael C. (2019): “When Speed Kills: Lethal Autonomous Weapon Systems, Deterrence and Stability”. Journal of Strategic Studies 42(6), pp. 764—788.
- Brose, Christian (2020): The Kill Chain: Defending America in the Future of High-Tech Warfare. Hachette Books.
- Bode, Ingvild / Watts, Tom (2023): Loitering Munitions and Unpredictability. Center for War Studies Research Report.