Drones and the Laws of War

Although the current laws of war can account for drones, technology may quickly outpace the law.

Unmanned military technology — of which Unmanned Aerial Systems (UAS – commonly called drones) are a large part — is governed by the same laws as a soldier with a gun. The same set of customary international law (CIL) and law of armed conflict (LOAC) principles

that apply to soldiers also apply to drones. CIL, in this context, involves the UN Charter Article 51, which preserves states’ rights to use self-defense when subject to “an armed attack.” The U.S. and others have generally agreed that the right of self-defense extends to anticipatory self-defense as well. The parameters of CIL self-defense in the context of the war on terror present a broader discussion than can be had here, but drones do highlight aspects of this debate by making possible tactical strikes that could not be carried out by conventional Special Forces.

LOAC refers to the principles of armed conflict codified, in part, in both the Hague and Geneva Conventions and Additional Protocols, also known as International Humanitarian Law. LOAC requires armed forces engaged in International Armed Conflict (IAC) or Non-International Armed Conflict (NIAC) to abide by proportionality and distinction. Proportionality prevents a state from using force that will result in too much collateral damage, and distinction requires states to use force only when civilians and combatants can be separated. In a statement at the American International Law Society on March 25, Harold Koh, the Legal Advisor to the U.S. Department of State, set out a legal defense of U.S. targeted-warfare and advanced weapons systems that was based on the principles of CIL self-defense and LOAC principles of proportionality and distinction. Thus, the U.S. currently treats drones as conventional weapons for legal purposes, and must abide by CIL and LOAC.

Although drones do fall under pre-existing law, the technology (and the evolution of warfare in general) requires a second look. Drones have been increasingly used to engage and destroy specific targets. In Pakistan, drone strikes increased from one strike in 2004 to 117 strikes in 2010 (data from The Long War Journal.Org). Drones have been used by non-military organizations, such as the CIA, to carry out strikes in a growing number of countries. Drones can arguably expand a non-military group’s (such as the CIA’s) capacity to carry out military-like actions, at very little cost in terms of casualties. Additionally, drones, combined with intelligence on the ground, are increasingly effective in delivering weapons systems. This raises questions about targeting, and an important aspect of LOAC – military necessity.

One side of the targeting and military necessity debate is that drones provide practical and precise options for using force. Using drones means the U.S., UN, or local troops and police forces do not have to engage hostile forces. Additionally, using drones can consolidate targeting decisions; ten soldiers necessarily provide ten different targeting decisions, ones with more concern about their security than a drone has – all of which increases adherence to military necessity. However, a counterargument to the use of drones is that they make the application of force too easy. The concept of military necessity, which stems from proportionality and distinction, needs to be revised with drone strikes in mind. Drones can reach into territory that U.S. ground forces cannot; this extended reach must be tempered with a refined definition of which targets are truly “indispensable for securing the complete submission of the enemy.”

Moving Forward:

An issue that remains somewhat on the back-burner, but will become exponentially more important as technology improves, is how autonomous drones can or should be. Currently, drones are operated by a single pilot by remote control. Like regular airplanes, much of the drone’s activity can be carried out by autopilot, so it is not too difficult to imagine a single pilot controlling a group of drones. It is also not difficult to imagine a drone programmed to react automatically in some instances, such as deploying defensive counter-measures or evading enemy fire. Autonomous weapons represent a large part of military R&D, and will soon necessitate legal distinction from conventional weapons, perhaps through an unmanned weapons convention or similar treaty.

Like a large portion of the current questions, the question of autonomy is focused on targeting. It will have to be decided whether a drone should be allowed to make its own targeting decisions, and if so, to what extent. This implicates the LOAC principles discussed above and creates a number of problems that would need to be addressed. For instance, can an autonomous weapon respond with distinction? Currently, no – for example, countries party to the Ottawa Treaty ban automatic anti-personnel mines because these weapons are incapable of distinction. Can autonomous weapons recognize surrender and take prisoners? Perhaps – for example, in the Persian Gulf War, a unit of the Republican Guard surrendered to a Pioneer drone from the USS Missouri. The most difficult question in autonomy will be one of responsibility. LOAC requires that a chain of responsibility extend from the use of force, and there is no clear answer on who would be responsible for a autonomous drone’s behavior.


Drones present a rapidly expanding military ability that, although currently accounted for in our laws of war, may soon outpace any current legal framework. To a large extent, drones merely highlight the evolution of warfare. However, technology is a proactive as it is reactive, and drones have the capability to drive change as much as follow it. Thus, as drones become less and less analogous to conventional weapons systems, the laws of war will need to keep up.