The skies over the conflict in Ukraine are filled with identical-looking consumer drones piloted by combatants and civilians alike – aircraft that promise to be a regular feature of war in the future. Few methods exist for telling small drones apart in airspace: combatants, unable to distinguish between drones that are a threat and those that aren’t, may direct attacks at civilian drone pilots.

In this post, civilian drone researcher and consultant Faine Greenwood argues that the aid world must move quickly to better define where small consumer drones fit into international humanitarian law.

Editor’s Note – The views expressed in this blog are those of the author and do not reflect the ICRC’s position or views.

Ukraine is engulfed in conflict, and small drones are flying over the battlefield. In a recent Facebook post, the Ukrainian Ministry of Defense requested that civilian drone-owners sign up to use their drones to assist military operations. A video uploaded to Telegram by an unknown user appears to show a Chinese-made DJI consumer drone capturing footage of a flaming mall in Chernihiv, while multiple individuals and charities are donating similar DJI drones to Ukrainian forces. One social media post appears to show captured Russian soldiers carrying DJI Mavic consumer drones with them into Ukraine, while another appears to be drone footage of Ukrainian citizens in Enerhodar, attempting to hold Russian tanks back. Journalists in Ukraine are using drones to document the scale of the destruction, including reporters from Reuters and Radio Free Europe.

Small consumer-intended and amateur-built drones aren’t a new phenomenon in the conflict. Since 2014, both Ukrainians and Russians have reportedly regularly incorporated consumer drones, often modified or hacked to override manufacturer restrictions on altitude, into their activities. These consumer drones often appear to be the same models journalists and OSCE monitors have been using to document the conflict for years – and the same models humanitarian aid workers outside of Ukraine and Russia now use widely in many places around the world.

Where do drones fit into international humanitarian law?

All this aerial activity raises an important question: how do small consumer and home-made drones, which are used by many different actors in conflict, fit into the Geneva Conventions and the Additional Protocols, and other treaty-based rules?

Many other uncertainties and questions flow from this central dilemma. For example: At what point does a consumer drone become a military objective as well as the person who’s piloting it? How can identical-looking consumer drones, which are usually too small to be visually identified in the air, unlike manned aircraft, be distinguished from one another in conflict? What are the implications of civilians from countries that aren’t parties to a given conflict donating consumer drones to one side’s armed forces, as is currently happening in Ukraine? And what might happen if a drone piloted by a humanitarian aid worker, a journalist, or a curious civilian is mixed up with a drone being flown by a combatant?

Some technologies available today make it possible to accurately pinpoint the location of both the drone itself and the drone pilot on the ground – who will be, in the case of small consumer drones, likely standing within a mile or less of their aircraft. Combatants might then decide to fire upon the pilot’s location, operating under the assumption that the drone is being flown by the enemy, not by a civilian. The regular presence of small, unidentifiable drones over conflict can also sow confusion, fear, and distrust: civilians on the ground will be unable to know who’s watching them, and for what purpose.

In 2022, it’s hard to argue that small consumer drones are military objects or objectives – no object is, until it meets both requirements of Article 52 of Additional Protocol I to the Geneva Conventions (see page 5, ILA Report on the Conduct of Hostilities). Globally, most civilians who own or build their own small drones use them for recreation, research, and work, not for military purposes: like other civilian technologies, they’re objects that can be adapted for military purposes, but weren’t explicitly designed for them.

That means there must be mechanisms for determining if a given consumer drone is a legitimate objective or not, and that’s where the IHL concepts of distinction, precautions, identification, and verification are key.  As recently stated by ICRC President Peter Maurer, the Geneva Conventions and the First Additional Protocol are fully applicable in the case of the international armed conflict in Ukraine.

The basic rule of distinction is codified in Article 48 of Additional Protocol I, which reads:

In order to ensure respect for and protection of the civilian population and civilian objects, the Parties to the conflict shall at all times distinguish between the civilian population and combatants and between civilian objects and military objectives and accordingly shall direct their operations only against military objectives.

This problem of distinction is also relevant to Article 52(3) of the 1977 Additional Protocol I, which states:

In case of doubt whether an object which is normally dedicated to civilian purposes, such as a place of worship, a house or other dwelling or a school, is being used to make an effective contribution to military action, it shall be presumed not to be so used.

This raises a key question: how might the ‘doubt’ rule apply to small consumer drones, and under what conditions might a small consumer drone qualify as a legitimate military target? Other documents in international humanitarian law shed some light on this issue, but also raise more questions.

Can drones be medical transports?

Consider the use of small drones to transport medical supplies, a practice that a number of companies are currently experimenting with around the world. Article 22 of the Fourth Geneva Convention states that aircraft ‘exclusively employed for the removal of wounded and sick civilians, the infirm and maternity cases, or for the transport of medical personnel and equipment, shall not be attacked, but shall be respected while flying at heights, times and on routes specifically agreed upon between all the Parties to the conflict concerned’. Article 22 also states that such aircraft ‘shall obey every summons to land’.

In the Additional Protocol to the Geneva Convention (API), Article 8 defines relevant terminology: this section defines ‘medical transportation’ as ‘the conveyance by land, water or air of the wounded, sick, shipwrecked, medical personnel, religious personnel, medical equipment or medical supplies protected by the Conventions and by this Protocol’.

Following from this, ‘medical transports’ are defined as ‘any means of transportation, whether military or civilian, permanent or temporary, assigned exclusively to medical transportation and under the control of a competent authority of a Party to the conflict’. Finally, ‘medical aircraft’ is defined as ‘any medical transports by air’.

Meanwhile, Article 24 of the Additional Protocol to the Geneva Convention (API) states that: ‘Medical aircraft shall be respected and protected, subject to the provisions of this part’. Articles 25 to 30 describe the conditions of that respect and protection. Article 27, which pertains to medical aircraft in areas controlled by an adverse Party to conflict, states in its first part that medical aircraft ‘shall continue to be protected while flying over land or sea areas physically controlled by the adverse Party provided that prior agreement to such flights has been obtained from the competent authority of the adverse Party’.

In its second part, it states that a medical aircraft that flies over areas controlled by the adverse Party either without a prior agreement or in deviation from that agreement’s terms must ‘make every effort to identity itself and to inform the adverse Party of the circumstances’. At the same time, the adverse Party must ‘make all reasonable efforts to give the order to land or to alight on water, referred to in Article 30, paragraph 1, or to take other measures to safeguard its own interests, and, in either case, to allow the aircraft time for compliance, before resorting to an attack against the aircraft’.

Might the respect and protection defined in Article 24 of the API then apply to these small medical drones? If we consider the definitions outlined in Article 8, the transportation of ‘medical equipment or medical supplies’ is an action that qualifies for protection, objects that small drones realistically can (and do) physically move from place to place. And might the requirements related to identification and to landing, outlined in Articles 27 and 22, also apply to small drones as well – expectations that will be very difficult for these objects to fulfill, for reasons elaborated upon later in this piece? The answer to both questions is unclear.

Another issue for small drones arises in Article 28 of the Additional Protocols, which describes restrictions on the operation of medical aircraft: it states that medical aircraft ‘shall not be used to collect or transmit intelligence data and shall not carry any equipment intended for such purposes’. Small drones, including those that are used for medical supply delivery, use a range of specialized sensors to chart their course through the air and to detect and avoid obstacles, including GPS receivers, barometers, acoustic sensors, and cameras. What’s more, standards for encrypting and securing this data vary very widely among small drone manufacturers and makers. It’s quite possible that a party to conflict could argue that these drone sensors are being used to collect and transmit intelligence data, or that the data they collect could be stolen or appropriated by the enemy.

The 2009 HPCR Manual on International Law Applicable to Air and Missile Warfare offers an authoritative restatement of existing international law applicable to air or missile operations in international armed conflict, and is a useful document for thinking through IHL and small drones. In the HPCR Manual, Rule 102 reiterates that humanitarian relief personnel must be respected and protected and that protection ‘extends to humanitarian transports, installations and goods’. Per the commentary, this rule is based upon Article 71(1) and (2) of the Additional Protocols. The commentary (which was drafted by select experts from the original group, and was not adopted by consensus) explicitly acknowledges the possibility of UAVs being used for humanitarian aid purposes, stating that aircraft ‘exclusively engaged in such activities pursuant to an agreement between the Belligerent Parties are entitled to specific protection from attack’.

Small drones are hard to spot

These rules all define cases in which civilian, medical, or humanitarian aircraft can and cannot be attacked. The next set of relevant rules describe how protected aircraft might distinguish themselves from others present in the airspace over conflict.

Article 18 of the Additional Protocols states that parties to conflict ‘shall endeavor to ensure that medical and religious personnel and medical units and transports are identifiable’, and that these parties should also ‘endeavor to adopt and to implement methods and procedures which will make it possible to recognize medical units and transports which use the distinctive emblem and distinctive signals’.  Article 22 of the Fourth Geneva Convention states that medical aircraft ‘may be marked with the distinctive emblem provided for in Article 38’ of the 1949 Geneva Convention.

Meanwhile, Rule 76 of the HPCR Manual says medical transports ‘ought to be clearly marked’ and that medical aircraft ‘ought to use additional means of identification when appropriate’. Rule 65 of the Manual, concerning humanitarian aid aircraft, states that protected aircraft can lose that protection if they don’t ‘comply with the details of the agreement, including availability for inspection and identification’. The commentary to the rule expands on this, noting that aircraft granted safe conduct are ‘obliged to comply with an order to identify themselves’.

These articles and rules, as they’re currently worded, present multiple problems for small drones.

To begin with, consumer and DIY drones are usually very small. People on the ground can realistically be expected to notice a red cross or another distinguishing emblem painted on the side of a relatively low-altitude manned helicopter or airplane: they can’t be expected to see a similar marker painted on the side of a 570 gram drone that’s moving at speed. While all drone pilots should be encouraged to mark their drones with appropriate symbols and information, these measures simply won’t be adequate to address the distinction problem.

Nor can drones, at least with the technology available on the battlefield in Ukraine today, readily identify themselves upon request. Unlike manned aircraft, small drones can’t be directly communicated with via radio: there’s no pilot in the cockpit who can respond to a radio demand for identification. The small consumer drones most likely to show up on modern battlefields don’t possess technology that might allow a message sent to the drone to be relayed to the pilot below. This also means that small drones will likely be unable to comply with a summons to land, as is described in Article 22 of the Fourth Geneva Convention.

While counter-drone and drone detecting systems do exist, like DJI’s mobile Aeroscope system they’re largely only able to report back that a drone is present, as well as the location of the drone and, sometimes, the location of the pilot. While some DJI products do give drone pilots the option of voluntarily broadcasting identifying information, it’s unclear how many people know about this, use it, or have the ability to receive it.

Beyond DJI’s proprietary system, other methods for remotely identifying drones in airspace do exist. Some companies sell ADS-B transponders that can be used for small drones, as well as IFF (Identification Friend or Foe) systems tiny enough to be used with a small drone. However, some countries, like the US, explicitly bar most drones from broadcasting ADS-B out signals, to avoid cluttering networks used for manned air traffic.

To address this network clutter issue, many countries around the world are in the midst of developing so-called Remote ID standards that will integrate small drones with existing air traffic control systems. But these systems are complex and very much still in-development: implementing such a system in the midst of the current conflict in Ukraine, with many other competing humanitarian priorities, would be an immense challenge.

Drones, distinction, and what to do about it

Addressing the problem of small drones and distinction in conflict, both in the context of the current Ukraine war and in the near future, is a formidable challenge. Here are some initial suggestions.

First and foremost: we need to raise awareness about the problem of consumer drones in conflict. All actors need to be made aware that they’re likely to encounter small drones in conflict settings: everyone should have basic information on their technical capabilities, what they look like, and the different ways in which they might be used. Combatants must be made aware that small drones (and their pilots) do not automatically constitute appropriate targets. Civilians, including reporters, must be made aware that using a small drone in conflict settings, in the absence of clear standards for distinction or identification, could put them at extremely serious risk.

Second, we need dialogue. Humanitarians, governments, journalists, drone-makers, and other interested parties should convene as soon as possible to discuss these specific issues around consumer drones in conflict, and how they might be better addressed. The Humanitarian UAV Code of Conduct, first developed in 2014 and regularly updated since, might present one centralized venue for compiling these best practices in a non-legally binding location.

One relatively low-hanging fruit issue that such a convening could tackle: defining widely-agreed upon standards for visually marking drones in airspace. Everyone who uses drones in conflict and disaster areas needs to be aware of existing rules on the use of the Red Cross emblem: they also need guidance on if there are situations where medical-delivery drones might be entitled to use it. Humanitarian drone users, journalists, and others who aren’t entitled to use the emblem should also consider developing standards for marking and identifying their own unmanned aircraft. Could lights or other methods be used in addition to visual markings (which will be, as mentioned earlier, very hard to see on their own)?

Another possible priority: defining and popularizing strategies for allowing civilian drone pilots to formally file advance flight plans with parties to conflict. This is reflected in the HPCR Manual, which notes in the commentary to Rule 64 that advance flight plans are one alternate method by which aircraft can identify themselves in conflict. It’s currently unclear how civilian drone pilots could (or should) file such plans. More clarity is needed.

These strategies will all need to be accompanied by technological solutions. The humanitarian world and other civilian organizations should work with governments and technology makers to come up with workable electronic solutions for identifying drones during war. As remote ID systems mature, decision makers should also contemplate how they might work during conflict and during other disasters, in dangerous and communication-restricted environments.

Time is of the essence. The conflict in Ukraine has made it clear that consumer drones are now a mainstream technology in war, not occasional oddities that can be safely overlooked. We need to act now to avoid the dangerous and confusing consequences of drone-filled skies over conflict zones.

See also

• Faine Greenwood, Drones and distrust in humanitarian aid, July 22, 2021
• Klaudia Klonowska, Shifting the narrative: not weapons, but technologies of warfare, January 20, 2022
• Damian Copeland & Lauren Sanders, Engaging with the industry: integrating IHL into new technologies in urban warfare, October 7, 2021
• ICRC, Handbook on data protection in humanitarian action, 2021