Chemical weapons and law enforcement under international law

ILPI Publications > Background Papers

An overview of the legal status of chemical weapons: prohibitions, obligations and exceptions.

December 2016


The use of chemical weapons is subject to comprehensive prohibitions under international law. Not only is the use of chemical weapons implicitly prohibited by the general rule of distinction under international humanitarian law (IHL), most imaginable uses of chemical weapons would contravene the proportionality rule under human rights law.



The use of chemical weapons is subject to comprehensive prohibitions under international law. Not only is the use of chemical weapons implicitly prohibited by the general rule of distinction under international humanitarian law (IHL), but most imaginable uses of chemical weapons would contravene the proportionality rule under human rights law.

In addition, the 1993 Chemical Weapons Convention (CWC) specifies that chemical weapons may “never under any circumstances” be used. However, the CWC contains an exception for instances where toxic chemicals are used for “law enforcement including domestic riot control purposes”.[1] In short, the legality of the use of chemical toxins depends on whether they are used as a means of warfare, or whether they are used for purposes not prohibited under the CWC, including law enforcement purposes. In the former situation (warfare), the use of toxic chemicals is always prohibited. In the latter (law enforcement), the use of toxic chemicals can, under certain circumstances, be allowed.

This paper considers two complex legal questions that the aforementioned exception generates. First, what is the relationship between “law enforcement” and “domestic riot control”? This issue is closely related to the question of which chemical agents may be used for law enforcement purposes including riot control. Second, how should we understand the term “law enforcement” as opposed to “acts of warfare”? This question similarly has important implications for whether or not toxic chemical agents may be used.

The 2002 Moscow Theatre incident, in which 126 of the more than 800 hostages were killed by the Russian authorities’ use of fenatyl gas, is a case in point illustrating some of these issues. Was the hostage taking an “extension” of the conflict in Chechnya? Was the use of gas an act of law enforcement or an act of warfare? If the situation could be classified as law enforcement, was the fenatyl gas allowed under the CWC exception?

This paper proceeds as follows. First, it provides a brief description of the functioning and types of chemical weapons. Second, it offers an overview of the rules of conventional and customary international law applicable to the use of toxic chemicals, including IHL, international criminal law and international human rights law, as well as the CWC itself. Third, it discusses the difficulties in ascertaining whether a situation amounts to warfare or law enforcement, given the apparent blurring of the lines between these situations in recent military operations. The paper then considers the law enforcement exception laid down in the CWC, and discusses the Moscow Theatre incident.

What are chemical weapons?

The scientific definition of a chemical is very wide; a chemical may be defined as any form of matter with a constant chemical composition and characteristic properties. The legal definition of a toxic chemical is also very wide, encompassing all chemicals that can cause permanent or temporary harm to humans or animals.[2] Chemical weapons are generally understood as “toxic synthetic chemicals that can be dispersed as a gas, liquid or aerosol or as agents absorbed to particles to become a powder.”[3] Modern chemical weapons are typically delivered by means of a missile or gravity bomb. Developing modern chemical weapons is fairly resource demanding. Manufacture of the most potent or dangerous gases, such as sarin, requires specialized equipment and know-how.[4]

Types and effects

A wide range of chemical agents have been used in warfare (see classification of types in the table below). A common effect of most chemical agents is suffocation and/or damage to the respiratory system.[5]

Definition of chemical weapons in relation to biological weapons

It is not always evident whether an agent should be classified as a chemical agent, and hence governed by the international instruments applicable to chemical weapons, or a biological agent, subject to the rules applicable to bioweapons. Presumably, a key difference between biological and chemical agents is that biological agents are made up of living organisms. Yet to define biological weapons as weaponized disease is increasingly inaccurate. With new technologies, biological organisms can be created from chemical substances or even informatics.


Chemical weapons under international law

Historical overview

Chemical weapons have a long record of use. Their international regulation goes back to Article 23 (a) of the 1899/1907 Hague Regulations.[12] Mustard gas – the “king of war gasses” – was used for the first time on 22 April 1916, at the 2nd battle of Ypres, during World War I. Negotiated in response to the horrors of the First World War, the 1925 Geneva Protocol prohibits the use of asphyxiating, poisonous, or other gases, and of all analogous liquids, materials or similar devices in war. This Protocol is generally hailed as the major explanation for the lack of (massive) use of chemical weapons as a means of warfare during World War II.

Article 3 of the Potsdam Agreement obligated Germany to distribute all war material to the Allies, or to destroy it. Dumping at sea was viewed as the safest disposal method at the time,[13] and at the Potsdam Conference, the Allies decided to dump the chemical munitions in the Atlantic Ocean. The agreement required the munitions to be transported by boat to the Atlantic Ocean and dumped at depths of more than 1000 metres. It appears that this requirement was widely disrespected. Chemical munitions were dumped in the North Sea and the Baltic Sea, sometimes in waters only tens of metres deep.[14] The locations of the dumped chemical weapons were classified for 50 years, a classification that was extended for 20 years with the entry into force of the CWC, and that will expire in 2017.

Chemical weapons have been repeatedly being used in the Syrian Civil War.[15] Their contemporary use is not limited to the Syrian conflict, however. In 1988, Saddam Hussain used mustard and sarin gas in the war against Iran, and against Iraqi Kurds in northern Iraq. Many thousand were killed. In 1995, the Japanese cult Aum Shinrikyo used sarin gas in a terrorist attack on the Tokyo subway. Thirteen people were killed and over 6000 injured.[16] In 2006 and 2007, al Qaeda in Iraq used chlorine in vehicle bombs deployed against US and Iraqi troops. Chemical artillery rounds in improvised explosive devices were also reported on several occasions.[17] The Taliban has reportedly also used chemical agents in attacks on schools in Afghanistan.[18]

Chemical weapons use and international humanitarian law

Most conceivable instances of chemical weapons use would breach with the general rules of IHL, because such use would be indiscriminate and possibly also cause superfluous injury or unnecessary suffering.

To the extent that chemical weapons are imprecise, and thus cannot be directed at specific military objectives alone, they are indiscriminate and therefore unlawful under Article 51(4)(b) of Additional Protocol I to the Geneva Conventions (API). When the effects of a chemical weapon may not be limited to military objectives alone, they may also be disproportionate, and thereby indiscriminate within the meaning of the protocol’s Article 51(4) (c). Certain chemical weapons may also cause unnecessary suffering and superfluous injury in enemy belligerents, and would therefore be prohibited according to AP I Article 35(2) as well as under international humanitarian customary law. There is also an explicit prohibition of toxic contamination of water-supply installations and foodstuffs in the Geneva Conventions and AP I.[19]

The use of chemical weapons is prohibited under Article 1(b) of the CWC. This prohibition includes the use of riot control agents as a method of warfare.[20] Unintentional and insignificant poisonous secondary effects of otherwise permissible munitions are excluded. Although six states remain outside of the CWC,[21] conventional and customary IHL prohibits the use of chemical weapons by any party to an armed conflict.[22]

Chemical weapons and international disarmament law

In addition to use, the CWC prohibits the development, production and stockpiling of chemical weapons. The Convention also obliges state parties to destroy of existing stockpiles and provides extensive verification and inspection regimes to be carried out by the OPCW.

An in-depth reflection on the future of the OPCW notes that the CWC implementation organisation “is the collective property and responsibility of the states parties but at the same time has become a global public good”.[23] According to the OPCW, 126 states parties have reported taking legislative and administrative measures to implement the CWC, while 182 “National Authorities” have been established to implement it.[24]

The CWC definition of a chemical weapon contains the so-called general purpose criterion whereby all toxic chemicals and their precursors are prohibited “except where intended for purposes not prohibited” and “as long as the types and quantities are consistent with such purposes”.[25] Unprohibited purposes are essentially those other than the use of toxic chemicals and their precursors as weapons. Thus, exceptions to the definition include “peaceful purposes”,[26] as well as, first, military purposes that do not involve the use of toxic chemicals as weapons,[27] and, second, law enforcement including domestic riot control.[28]

The general purpose criterion is what makes the CWC’s prohibition against chemical warfare comprehensive in scope, and thus takes potential future changes in science and technology into account. States Parties have agreed to a verification regime for certain toxic chemicals and their precursors (listed in Schedules 1, 2 and 3 in the CWC Annex on Chemicals)[29] in order to ensure that such chemicals are only used for unprohibited purposes. Those toxic chemicals must be declared by state parties.

A CWC state party is not required to declare chlorine or destroy it under the CWC because of its many civilian purposes. However, any use of chlorine or any other toxic chemical in warfare is illegal under the convention. The UN Security Council has through Resolution 1540 (2004) required states to implement national measures to prevent non-state actors from using chemical and biological weapons.[30]

Chemical weapons and international criminal law

The Rome Statute makes it a war crime to employ “poison or poisoned weapons”. Its Article 8 (2)(xvii)  refers to “employing asphyxiating poisonous or other gases, and all analogous liquids, materials or devices”. These expressions are generally understood to cover chemical weapons.[31] However, it is sometimes suggested that the terms “poisons” and “poisonous gases” might exclude at least some types of chemical weapons. Some interpret the Rome Statute, which includes these two terms, to exclude both chemical and biological weapons.[32] Further, the Statute’s negotiating history shows that delegates refused to include explicit references to biological and chemical weapons. This has led some commentators to conclude that the Statute must be interpreted in a manner that excludes chemical and biological weapons from its scope.[33] A less extreme view posits that only certain kinds of chemical weapons are excluded. Though the passage “asphyxiating, poisonous or other gases, and of all analogous liquids, materials or devices” has widely been interpreted to include some chemical weapons, it might not cover other chemical agents since they are not necessarily “poisons”. Regardless of which interpretation is correct, this lack of consensus reflects a need for clarification.[34]

The CWC requires that state parties make all acts prohibited under the Convention also unlawful for “natural and legal persons anywhere on its territory or in any other place under its jurisdiction”.[35] International criminal law, however, only pertains to natural persons.[36] The CWC does not clarify what legal implications the use of toxic chemicals by a non-state group in a conflict location will have, or how the convention’s general-purpose criterion can be operationalized. Individual members of a non-state group employing chemical weapons as a means of warfare could be prosecuted by the International Criminal Court (ICC), depending on the interpretation of Article 8 (2)(xvii) of the Rome Statute as discussed above.

The Chemical Weapons Convention and law enforcement


As noted, the CWC prohibits use of toxic chemicals in all circumstances, not only in armed conflict.[37] One exception from this prohibition[38] is for law enforcement including domestic riot control purposes.[39]

Defining the law enforcement exceptions was contentious during the CWC negotiations. This remains a difficult and disputed issue, both with regard to classification of the de facto situation (do law enforcement rules apply or does the situation amount to an armed conflict?), and with regard to what kind of chemicals that can legally be employed in specific situations.

The distinction between hostilities and law enforcement is not always clear. Outside of armed conflict, several international legal frameworks regulate the use of toxic chemicals against persons. They include the CWC and the BWC, as well as international human rights law which regulate use of toxic chemicals for law enforcement. Nevertheless, diverging views have risen about the scope of the law enforcement exceptions in the CWC.

The meaning of “law enforcement including domestic riot control purposes”

The point of departure is therefore that toxic chemicals (and their precursors) may not be used against people, save for a very limited range of chemicals and in a very limited range of situations. These exceptions pertain to law enforcement. Chemical weapons are always banned as a means of warfare.[40] It follows that their use is never lawful in the context of an armed conflict.[41]

Law enforcement implies that the state concerned has jurisdiction. A state cannot enforce laws unless it has jurisdiction or authority to do so. However, this does not necessarily mean that law enforcement may only take place on a state’s national territory. A state can have several legal bases for law enforcement operations abroad, such as consent of the host state or authorisation by the UN Security Council.

Riot control is a narrow concept.[42] The CWC specifies that only Riot Control Agents (RCAs) can be used for riot control purposes. RCA are chemicals which can “produce rapidly in humans sensory irritation of disabling physical effects which disappear within a short time following termination of exposure.”[43] These are typically chemicals known as tear gas or pepper spray.

The CWC contains no definition of the expression “law enforcement including domestic riot control”, but the wording of Article II(9)(d) indicates that “law enforcement” has a broader meaning than “domestic riot control”. Riot control is a sub category of law enforcement, hence the word including. According to the text of the CWC, therefore, law enforcement is a term that is wider than “domestic riot control”.[44]

The prohibition against use of chemical toxins does not refer explicitly to the the exception of “law enforcement including domestic riot control”. A key question is therefore what kind of toxic chemicals might be allowed under this exception. Several commentators argue that only RCAs can be used for law enforcement under the exception in Article II(9)(d).[45] This assertion is doubtful. First, the very definition of a toxic chemical in Article II(1)(a) refers to all of the “purposes not prohibited under this Convention”. These purposes include law enforcement. A toxic chemical is defined as any chemical that can cause death, temporary incapacitation or permanent harm to humans or animals.[46] RCAs are chemicals that can cause temporary incapacitation and therefore fall under the broader definition of toxic chemicals. Because the definition of a toxic chemical also points to the exceptions for law enforcement, it seems that the CWC foresees the possibility of chemicals other than RCAs being used for law enforcement as well.

Moreover, the CWC Verification Annex treats the chemicals listed as Schedule 1, 2 and 3 chemicals in a manner that indicates that the law enforcement exception is not limited to RCAs. Part VI of this Annex explicitly states that “[a] State Party shall not produce, acquire, retain, transfer or use Schedule 1 chemicals unless the chemicals are applied to research, medical, pharmaceutical or protective purposes”.[47] Thus, Schedule 1 chemicals may never be used for law enforcement. Fidler points out that, since Schedule 2 and 3 chemicals do not have such restrictive language, the Convention implies that these toxic chemicals can be used for law enforcement purposes.[48] In other words, the Verification Annex also envisages the potential use of toxic chemicals other than RCAs for law enforcement.

In addition, those subscribing to the view that the CWC dictates that only RCAs may be used for law enforcement offer no cogent authority in their argument for this interpretation.[49] The three sources cited by Krutzsch and Trapp are: The Finogenov et al. case before the European Court of Human Rights, a case that did not find the use of incapacitating gases per se to be in breach of the European Convention on Human Rights (see below); the 1925 Geneva Protocol, from which Article XIII of the CWC prohibits any interpretive deviations; and the “broad consensus” shown in the literature that the Geneva Protocol bans the use of all chemical agents in war. None of these materials excludes the use of toxic chemical others than RCAs for law enforcement purposes as such.  Krutzsch and Trapp assert that “[a]ccepting, in principle, the use of certain toxic chemicals such as ICAs [incapacitating chemical agents] for law enforcement in exceptional situations, for instance during an antiterrorist operation, would undermine the prohibitions of Article 1 of the Convention and wrongly imply that there was a ‘choice’ between applying binding rules of treaty interpretation and replacing them by explanations constructed as required by military interest”.[50] This seems to be a statement of law as it should be.

Significantly, during the negotiations of the CWC, the United States argued for a formulation that would allow it to continue carrying out capital punishment with toxic chemicals. The US legislation implementing the CWC into national law explicitly defines law enforcement purposes as “[a]ny law enforcement purpose, including any domestic riot control purpose and including imposition of capital punishment[51] (emphasis added). Countries where the death penalty is practiced negotiated to keep the Convention deliberately open so that Article II (9) (d) could be read to allow them to use toxic chemicals which are not RCAs for capital punishment.[52]

The usage of toxic chemicals other than RCAs in law enforcement has also been discussed in the context of other international legal frameworks. In 2013, the ICRC adopted a position to the effect that toxic chemicals as weapons for law enforcement purposes should be limited to RCAs.[53] According to this view, the CWC, the BWC, international human rights law and international drug control law form an “overlapping legal framework regulating the use of toxic chemicals for law enforcement”.[54] The ICRC moreover stated that, “this legal framework leaves little room, if any, for the legitimate use of toxic chemicals as weapons for law enforcement other than the use of riot control agents, as defined in the Chemical Weapons Convention”. This seems to be a reasonable description of international law as a whole. The point here is merely that the CWC per se cannot be interpreted as excluding the use of toxic chemicals other than RCAs for law enforcement purposes.[55]

The distinction between “law enforcement” and “method of warfare”

Article I (5) of the CWC specifically prohibits the use of RCAs as a method of warfare. One reason for including this provision in the CWC, was to prevent situations such as those during the Vietnam war, when thousands of people died because of the use tear gas on the battlefield. The Commentary to the CWC explains that this provision was added to Article I at the very end of the negotiations: “It was a reaction to concerns raised mostly by non-aligned developing countries that riot control agents might be considered permitted for use by military forces.”[56] CS – a tear gas – had been used for “terrain denial” along the Ho Chi Minh trail, to “flush out” combatants from cover or from bunkers, and as a “force multiplier” to disable targeted persons to more effectively expose them to lethal force.[57]

Since RCAs fall within the definition of a chemical weapon, the specification that they cannot be used in warfare was not strictly necessary. Any use of any chemical weapon is prohibited in combat, and the only exception regarding toxic chemicals pertains to law enforcement. The difficulty is that it is not always clear whether law enforcement rules or combat rules apply. The question of exactly when a situation moves from unrest, disturbances or riots to armed conflict is not easy to answer, particularly in contexts approaching the lower threshold of armed conflict. Clear examples of each situation are not hard to find, but it is equally easy to find “grey zone” cases where a situation is moving back and forth between low-intensity armed conflict and law enforcement. For example, during their engagement in Afghanistan, the International Security Assistance Force (ISAF) soldiers switched back and forth between practicing law enforcement and engaging in combat operations. There has been growing concern over what is often called “blurring of the lines” between law enforcement and combat in recent international operations. The counterinsurgency (COIN) strategy adopted by ISAF commanders in Afghanistan following the “surge” in 2007 has made this distinction even more blurred.[58] It is not clear whether COIN is law enforcement or combat.

Through an executive order, the United States has interpreted the law enforcement exception to include (a) riot control of prisoners of war, (b) situations where civilians are used to mask attacks, (c) rescue missions and escaping prisoners, and (d) protection of military convoys from “civil disturbances, terrorists and paramilitary organisations”.[59] It would appear that, depending on the specific situations, activities described in (b) through (d) could be carried out in violation of the prohibition against using chemicals in combat.

Where it is unclear whether law enforcement rules or combat rules apply, soldiers on the ground carry the risk of acting in violation of the law as determined in national or international courts. Should a court decide that the situation must be classified as armed conflict, then the use in hostilities of toxic chemicals, even RCAs, could constitute a war crime.

Legality of the use of fentanyl derivatives in the Dubrovka hostage crisis under the Chemical Weapons Convention


On 23 October 2002, a group of armed men and women dressed in plain clothes arrived at the Dubrovka Theatre in Moscow. Minutes later, gunmen stepped on stage during a musical, forcing the actors to join the spectators and taking approximately 900 people hostage. The terrorists were a group of Chechen rebels who demanded that Russian forces be withdrawn from Chechnya and the region given independence.[60] The terrorists also declared that they would blow up the theatre if Russian forces attempted a rescue mission. Over the next three days, the terrorists released over 200 hostages, mainly children and Chechen nationals who were able to produce proof of their nationality. The negotiations did not proceed smoothly, however, as the terrorists threatened several times to begin executing hostages if their demands were not met.[61]

Before attacking the theatre, the Russian military released a gas into the theatre’s air conditioning system. The release of the gas caused 126 hostages to die from gas poisoning. While several died on the spot, most died as a result of delayed medical treatment. Russian authorities did not promptly give ambulances access to the scene, or facilitate the transport of wounded hostages to hospitals. Moreover, the authorities did not release the identity of the gas to the doctors, which resulted in many hostages dying before the doctors could find the proper antidote.[62]

To this day, the exact nature of the gas used remains uncertain. Russian authorities officially claimed that the gas was fentanyl. The antidote to fentanyl did not work in all cases, however, and traces of other elements were found in the bodies of foreign nationals. This has led experts to conclude that the gas was a combination of aerosolized fentanyl derivatives and an inhalational anaesthetic.[63]

The families of foreign nationals that were killed by the gas brought a suit against Russia in the European Court of Human Rights (Finogenov et al. v. Russia). As noted earlier, the petitioners did not contest Russia’s actions under the CWC. Rather, their complaint relied on Article 2 of the European Convention on Human Rights that safeguards the right to life. The Court found Russia to be in violation of the European Convention with regard to the right to life because of the lack of preparedness to treat those affected by the gas.[64]

The question is: was Russia’s use of the fentanyl derivatives in compliance with international law beyond human rights law, particularly the CWC?

The classification of fentanyl

As noted above, the CWC defines an RCA as any chemical not listed in its Schedule that can produce rapidly in humans sensory irritation or disabling physical effects which disappear within a short time following termination of exposure.[65] While fentanyl is not listed in any schedule, it still does not meet the definition of an RCA because its effects do not disappear without medical intervention or the administration of an antidote.[66] Another distinction comes from the fact that all currently recognized RCAs target the peripheral nervous system, while fentanyl targets the central nervous system.[67] Thus, fentanyl is not an RCA but a substance known as “incapacitating chemical agent” (ICA).[68]

Are there any restrictions on using ICAs in law enforcement according to the CWC?

The CWC does not explicitly allow ICAs to be used in law enforcement (it does not even mention ICAs), but the most logical reading of the Convention is that it also does not explicitly prohibit such use.  The Convention does, however, limit use to “the types and quantities [that] are consistent with such purposes.” [69]

The “type” requirement poses the greatest difficulty to reconciling the use of fentanyl derivatives with the CWC criteria.[70] Neill postulates that a chemical might not be considered appropriate for law enforcement purposes if its biological activity cannot be reversed.[71] After all, the purpose of law enforcement is rarely to harm or kill persons.

What matters here is not just the availability of an antidote in general, but its tactical capacity of being administered soon after exposure to the toxic chemical. Additionally, problems arise if the difference between the chemical’s incapacitating dose and lethal dose is too small for the former amount to be administered with a reasonable degree of confidence while avoiding the latter.[72]

Regarding the implied “quantity requirement”, the means of delivering an ICA can help determine whether its use violates the CWC. The Convention’s commentary observes that the means must be designed to enable law enforcement officers to use them safely for the permitted purposes.[73] Small quantities of an ICA with hand-held delivery systems might be suitable for law enforcement, but large quantities with large delivery systems would begin to appear akin to combat systems.[74]

While Russia’s delivery system certainly indicated a typical law enforcement situation – a fentanyl derivative was released into the air conditioning system rather than delivered by means of a mortar – Russia may have nevertheless have violated the CWC because the fentanyl derivate was not of the “type” that was safe for law enforcement.

Are there any restrictions on using ICAs in law enforcement according to the European Convention on Human Rights?

In Finogenov, Russia was accused of violating, inter alia, the right to life through the use of fenatyl gas. The Court found that the Convention had indeed been violated, but only with regard to the lack of medical preparedness in the aftermath of the gas’ use and the storming of the theatre.

The right to life is laid down in Article 2 of the Convention. It specifies that everyone’s right to life shall be protected by law, and that any use of force resulting in loss of life is a violation of the provision if the use of force was “more than absolutely necessary”.[75] The Court stated that it

acknowledges that in such situations some measure of disorder is unavoidable. It also recognizes the need to keep certain aspects of security operations secret. However, in the circumstances the rescue operation of 26 October 2002 was not sufficiently prepared, in particular because of the inadequate information exchange between various services, the belated start of the evacuation, limited on-the-field coordination of various services, lack of appropriate medical treatment and equipment on the spot, and inadequate logistics. The Court concludes that the State breached its positive obligations under Article 2 of the Convention.[76]

Russia had failed to provide prompt medical assistance, resulting in many if not most of the deaths. There was no medical assistance on the buses used for transportation of victims to hospitals. Nor was there any plan for the distribution of victims amongst various hospitals. All this led to delays in the provision of medical care for the victims, with some victims left lying in front of the theatre for over an hour without appropriate medical treatment.[77]

The Court found, however, that there had not been a violation of Article 2 with regard to “the decision by the authorities to resolve the hostage crisis by force and to use the gas”.[78] The question still remains whether it was at all possible to provide adequate assistance in such circumstances. One participant in an ICRC expert meeting noted that the expectation of providing adequate medical treatment is unrealistic in fast-moving tactical situations and in view of all the uncontrollable variables.[79] In addition, at least with respect to the specific chemical compound used, the difference between the lethal and the incapacitating doses was too small to render the compound consistent with the purposes of law enforcement.[80] More generally, Lynn Klotz, Martin Furmanski and Mark Wheelis calculated the “therapeutic index” of many possible ICAs and concluded that no current pharmaceutical product offers a high enough “safety margin”.[81]

It is possible that Russia was not found to have violated Article 2 of the European Convention because this was the first instance of an ICA being used on such a large scale and the government may have genuinely believed the gas to be safe at the time. The Court found a violation due to lack of proper planning, but no violation due to miscalculating the dosage of the fentanyl derivative.[82] Now that the toxic effects of fentanyl derivatives are known, the legal landscape may be different.

Was this a law enforcement operation?

Russia would be in breach of the CWC should the Dubrovka Theatre incident form part of a non-international armed conflict. If this were the case, neither the type nor the amount of chemicals used would be relevant because the CWC bans the use of any chemicals in warfare – a term that covers both international and non-international armed conflicts.[83]

There are three threshold questions. First, did the Dubrovka incident itself constitute an armed conflict between the Russian government and the hostage-takers? Second, if the incident itself did not constitute an armed conflict, was there an armed conflict between the Russian government and the Chechen rebels at the time? Third, if such a conflict existed, was the Dubrovka incident part of it?

In its 1995 Tadić Jurisdiction Decision, the Appeals Chamber of the International Criminal Tribunal for the Former Yugoslavia (ICTY) held that a non-armed conflict exists wherever there is “protracted armed violence between governmental authorities and organized armed groups … within a State”.[84] According to the ICTY’s post-Tadić jurisprudence on the matter, this definition entails two criteria:  (1) the level of intensity of the armed violence and (2) the degree of the armed group’s organisation. Over time, ICTY judges have compiled the following list of indicators, though they are by no means mandatory or exhaustive:

Level of intensity:[85]

  • The seriousness of attacks and whether there has been an increase in armed clashes;
  • The spread of clashes over territory and over a period of time;
  • Any increase in the number of government forces and mobilisation and the distribution of weapons among both parties to the conflict;
  • Whether the conflict has attracted the attention of the United Nations Security Council, and whether any resolutions on the matter have been passed;
  • The number of civilians forced to flee from the combat zones;
  • The type of weapons used, in particular the use of heavy weapons, and other military equipment, such as tanks and other heavy vehicles;
  • The blocking or besieging of towns and the heavy shelling of these towns;
  • The extent of destruction and the number of casualties caused by shelling or fighting;
  • The quantity of troops and units deployed;
  • Existence and change of front lines between the parties;
  • The occupation of territory, and towns and villages;
  • The deployment of government forces to the crisis area;
  • The closure of roads; and
  • Ceasefire orders and agreements, and the attempts of representatives from international organisations to broker and enforce ceasefire agreements.

Degree of organisation:[86]

  • Factors signalling the presence of a command structure;
  • Factors indicating that an armed group could carry out operations in an organised manner;
  • Factors indicating a level of logistics have been taken into account;
  • Factors relevant to determining whether an armed group possessed a level of discipline and the ability to implement the basic obligations of Common Article 3; and
  • Factors indicating that the armed group was able to speak with one voice.

Commentators widely accept the twin intensity-organisation criteria, as well as their respective indicia listed above, as authoritative.[87]

Based on the foregoing, one may safely conclude that the Dubrovka incident itself is unlikely to have amounted to a non-international armed conflict involving the Russian security forces and the hostage-takers.

The next question, then, is whether there was a non-international armed conflict in October 2002 between the Russian government and Chechen rebels elsewhere. The Tadić Jurisdiction Decision contains observations about the temporal and territorial scope of IHL’s applicability to non-international armed conflicts that are of relevance to our discussion here. According to the Appeals Chamber, IHL applies to a non-international armed conflict “until … a peaceful settlement is achieved”.[88]

This description finds widespread support in the literature.[89] In its 2016 commentary on Geneva Convention I, the ICRC lists a number of considerations in order to determine when “a peaceful settlement is achieved”. They include, inter alia: (a) a complete military defeat a party, its demobilisation or a dissolution; (b) dissipation of armed confrontations without formal ceasefire; and (c) a lasting cessation of armed confrontations.[90] The commentary also adds as indicative elements the effective implementation of a peace agreement or ceasefire; declaration by the parties that they renounce all violence; the dismantling of government special units created for the conflict; the implementation of disarmament, demobilisation and/or reintegration programmes; the increasing duration of the period without hostilities; and the lifting of a state of emergency or other restrictive measures.[91]

In the early 1990s, following the collapse of the Soviet Union, separatists in the newly formed Russian Federation Republic of Chechnya started an independence movement called the Chechen All-National Congress. From 1994 to 1996, the Russian government fought the Chechens in a conflict that became known as the First Chechen War. Tens of thousands of civilians died, but Russia failed to win control of Chechnya’s mountainous terrain, giving Chechnya de facto independence. In August 1999, Chechen militants invaded the neighbouring Russian republic of Dagestan to support a local separatist movement. The following month, five bombs exploded in Russia over a ten-day period, killing almost three hundred people. Moscow blamed Chechen rebels for the explosions, which comprised the largest coordinated terrorist attack in Russian history. The Dagestan invasion and the Russian bombings prompted Russian forces to launch the Second Chechen War. In February 2000, Russia recaptured the Chechen capital of Grozny. The 2001 Battle for Vedeno marked the end of major military confrontations in Chechnya. Russia officially declared that the conflict had ended in April 2002.[92] Since then, the rebels have mostly engaged in guerrilla and terrorist acts, targeting Russia’s civilian population.[93]

It is therefore difficult to say, with any degree of certainty, whether a non-international armed conflict between the Russian authorities and Chechen rebels existed in October 2002. Even if it did, however, it remains doubtful whether the Dubrovka incident would have formed part of that conflict.

According to the Tadić Jurisdiction Decision, the law applies in “the whole territory under the control of a party, whether or not actual combat takes place there”.[94] This part of the Tadić definition is also widely accepted.[95] While IHL’s applicability to the whole of the territory of a state may be conceded, the ICRC commentary cautions that “it does not mean that all acts within that territory therefore fall necessarily under the humanitarian law regime” and that its application “requires a certain nexus”.[96] The commentary continues:

Furthermore, if a specific act carried out or taking effect in more peaceful areas of a State could … generally fall under the scope of application of humanitarian law, questions regarding the applicable legal standards in a particular scenario might still arise. It may also need to be determined whether, in a given case, a specific use of force is necessarily governed by conduct of hostilities law or whether it is governed by the law enforcement regime based on human rights law.[97]

The situation becomes “less clear … with regard to the use of force against isolated individuals who would normally be considered lawful targets, under international humanitarian law but who are located in regions under the State’s firm and stable control, where no hostilities are taking place and it is not reasonably foreseeable that the adversary could readily receive reinforcement”.[98]

In October 2002, the city of Moscow was firmly under state control, no hostilities were taking place there, and it was not reasonably foreseeable that the hostage-takers would receive reinforcement. This indicates that the Dubrovka incident was hardly governed by IHL’s rules regulating the conduct of hostilities.


The legal regulation of toxic chemicals in law enforcement is complex because the issue is determined not only by several legal frameworks, but also by the nature of the circumstances «on the ground», that is, whether a situation should be classified as «law enforcement» or «armed conflict». The question of how to classify a situation is crucial: If the threshold for an armed conflict has been reached, toxic chemicals can never be used in combat. It remains doubtful, however, whether the Moscow theatre incident in 2002 constituted armed hostilities as part of a non-international armed conflict.

When a situation is below the threshold of an armed conflict, certain toxic chemicals can be used in law enforcement, but only in very limited circumstances. There is a distinction between the narrow concept of “domestic riot control” and “law enforcement purposes” with regard to what kind of toxic chemicals may be lawful in specific situations.

Although incapacitating chemical agents (ICAs) cannot be classified as riot control agents (RCAs), their use is not categorically forbidden in law enforcement under the CWC. Still, it may be argued that in the case of the Moscow theatre incident in 2002, the type of incapacitating agent used and its small safety margin was not consistent with the purpose of law enforcement under the CWC. Moreover, use of ICAs in law enforcement may be unlawful under other legal regimes, because, as some commentators argue, no incapacitating agents currently exist that are sufficiently safe to be used in law enforcement operations.

* Gro Nystuen, Kristina Alekseyeva, Nobuo Hayashi, Cecilie Hellestveit and Magnus Løvold.


[1] According to Article II (1) (a), toxic chemicals are not considered chemical weapons when intended for “purposes not prohibited under this Convention”. Law enforcement is one of these purposes. Accordingly, a tear gas grenade, for example, is not considered a chemical weapon under the convention.

[2] Article II (1)-(4) of the CWC.

[3] K. Ganesan, S.K. Raza and R. Vijayaraghavan, “Chemical Warfare Agents”, Journal of Pharmacy and Bioallied Sciences, vol. 2, no. 3 (July–September 2010), p. 166.

[4] Federation of American Scientists, “Chemical Weapons Production and Storage”, undated. Available from:

[5] OPCW, “What is a Chemical Weapon?”, Fact Sheet no. 4, (accessed 16 December 2016).

[6] Kjølv Egeland, “Longing for Armageddon”, International Law and Policy Institute, WMD Project, Background Paper no. 16, 2015, (accessed 16 December 2016).

[7] Crowley, Michael and Dando, Malcolm, Submission by Bradford Nonlethal Weapons Research Project to Foreign Affairs Select Committee Inquiry on Global Security: Non-Proliferation, 2008, pp. 12, available at:

[8] Tsuji, et al, “Research Strategies for Safety Evaluation of Nanomaterials, Part IV: Risk Assessment of Nanoparticles” (2006) 89(1) Toxicological Sciences 42.

 [9] Guo, Peixuan “RNA Nanotechnology: Engineering, Assembly and Applications in Detection, Gene Delivery and Therapy” (2005) 5(12) Journal of Nanoscience and Nanotechnology 1964.

[10] See generally, Peter Dombrowski and Eugene Gholz, ‘Buying Military Transformation: Technological Innovation and the Defense Industry’ (2006); Henry C Bartlett, et al, ‘Force Planning, Military Revolutions and the Tyranny of Technology’ (Fall 1996) 24(4) Strategic Review 28; Nanotechnology and the International Law of Weaponry, p. 24.

[11] M. Wheelis and M. Dando, ibid., p. 560.

[12] Convention (IV) respecting the Laws and Customs of War on Land and its annex: Regulations concerning the Laws and Customs of War on Land. The Hague, October 1907.

[13] Bowles, R., 2006. Beaufort’s Dyke background – DSC-Env1, 21 August 2006, Ministry of Defence,

United Kingdom. 2006. BA5AC8EFB45C/0/bd_backgrd.doc

[14] The issue of sea-dumped chemical weapons is receiving increased attention, but the legal responsibilities for these munitions is limited. These chemical munitions are explicitly excluded from the scope of the CWC. Munitions dumped after WWII are explicitly not included in the verification mechanisms in the CWC, Article III.2 and IV.17: “The provisions of this Article and the relevant provisions of Part IV of the Verification Annex shall not, at the discretion of a State Party, apply to chemical weapons buried on its territory before 1 January 1977 and which remain buried, or which had been dumped at sea before 1 January 1985.”

[15] See inter alia 7th Report of the independent international commission of inquiry on the Syrian Arab Republic, UNHRC, 12 February 2014, A/HRC/25/65 paras 127 et seq (sarin) ; 8th Report of the independent international commission of inquiry on the Syrian Arab Republic, 13 August 2014, A/HRC/27/60, paras 115-118 (chlorine).

[16] Kjølv Egeland, “Longing for Armageddon”, International Law and Policy Institute, WMD Project, Background Paper no. 16, 2015, (accessed 16 December 2016).

[17] Iraq: The Fear Factor in Chlorine Bombs, 19, 2007. Stratfor, Global Intelligence.

[18] Kjølv Egeland, “Longing for Armageddon”, International Law and Policy Institute, WMD Project, Background Paper no. 16, 2015, (accessed 16 December 2016).

[19] See API article 54;  APII article 14.

[20] Article I.7. Each State Party undertakes not to use riot control agents as a method of warfare.

[21] Israel and Myanmar are signatories, while Angola, Egypt, North Korea and South Sudan are non-members.

[22] ICRC Customary Study rule 74 with references, Jean-Marie Henckaerts and Louise Doswald-Beck (eds.), Customary International Humanitarian Law, Vol. I: Rules.

[23] Note by the Director General, Report of the Advisory Panel on Future Priorities of the Organisation for the Prohibition of Chemical Weapons, Technical Secretariat, document S/951/2011, 25 July 2011, p. 3, para. 2.

[24] Organisation for the Prohibition of Chemical Weapons, CWC National Authorities, available at

[25] Article II. (1).

[26] Article II (9) (a) of the CWC.

[27] Article II (9) (b)-(c) of the CWC.

[28] Article II (9) (d) of the CWC.


[30] Security Council Resolution 1540 (2004) [Non-proliferation of weapons of mass destruction], 28 April 2004, S/RES/1540 (2004): ‘2. Decides also that all States, in accordance with their national procedures , shall adopt and enforce appropriate effective laws which prohibit any non- State actor to manufacture, acquire, possess, develop, transport, transfer or use nuclear, chemical or biological weapons and their means of delivery, in particular for terrorist purposes, as well as attempts to engage in any of the foregoing activities, participate in them as an accomplice, assist or finance them.’

[31] Jean-Marie Henckaerts and Louise Doswald-Beck (eds.), Customary International Humanitarian Law, Vol. I: Rules. International Committee of the Red Cross. Cambridge, Cambridge University Press, 2005. pp. 251-254; see Anders Boserup, The Problem of Chemical and Biological Warfare, Vol. III: CBW and the Law of War. Stockholm International Peace Research Institute. New York, Humanities Press, 1971. pp. 93-96.

[32] Cottier, 415. Some delegates at the Rome Conference were unaware that this language was understood to include chemical weapons.

[33] Gerhard Werle, Principles of International Criminal Law. The Hague, T.M.C. Asser Press, 2005. p. 371.

[34] Boserup, p. 41.

[35] Article VII (1) (a) CWC.

[36] ICC, Article 25 (1).

[37] Article I (1) (b).

[38] As mentioned above, the exception extends to a number of peaceful purposes, see Article II (9).

[39] Article II (9) (d).

[40] Article I (1).

[41] This does not preclude the lawful use of chemical weapons during an armed conflict as a measure of law enforcement.

[42] There is no definition of ”law enforcement including domestic riot control” in the CWC.

[43] Article II (7).

[44] Article 32 of the Vienna Convention of the Law on Treaties stipulate that a treaty must be interpreted ”in good faith” and ”in accordance with the ordinary meaning to be given to the terms of the treaty in light of its object and purpose.” Given that the object and purpose of this treaty encompass the prohibitions, obligations and exceptions alike, there is little guidance in this rule as to how to interpret Article II (9) (d). The ordinary meaning of the terms of the treaty on the other hand, entails that the term law enforcement includes more than domestic riot control.

[45] See e.g. Krutzsch, Myjer, Trapp (eds.) in The Chemical Weapons Convention – A Commentary, Oxford University Press 2014, page 94-95.

[46] Article II (2) of the CWC.

[47] VI.A.2(a), Verification Annex to the CWC.

[48] Fidler, D (2005) The meaning of Moscow: “Non-lethal” weapons and international law in the early 21st century. International Review of the Red Cross, Vol 87, No 859, September 2005, p 536.

[49] Krutzsch and Trapp, pp. 100-101.

[50] Ibid., p. 102.

[51] See Chemical Weapons Convention Implementation Act of 1998, SECTION 3, DEFINITIONS.

[52] Neill, D (2007) Riot control and incapacitating chemical agents under the chemical weapons convention. Defence R&D Canada Centre for Operational Research and Analysis, June 2007, p 7.

[53] ICRC position on the use of toxic chemicals as weapons for law enforcement, Policy 06.02.2013.

[54] Ibid. See also Michael Crowely, Chemical Control: Regulation of Incapacitating Chemical Agent Weapons, Riot Control Agents and their Means of delivery, Palgrave Macmillan, Basingstoke, 2016, chapter 11.

[55] See below for the possible use of toxic chemicals other than RCAs for law enforcement purposes in the context of the 2002 Moscow theatre tragedy where 125 people were killed by fenatyl gas.

[56] Krutzsch, W (2014) Article I: General Obligations, in: Krutzsch, W, Myjer, E and Trapp, R (eds.) The Chemical Weapons Convention: A Commentary Oxford University Press, p 69.

[57] ICRC (2010) Report of an Expert Meeting. “Incapacitating chemical agents”: Implications for international law, Montreux, Switzerland, 24-26 March 2010, p 18,

[58] On COIN, see for example the McChrystal strategy from 2009:

[59] Executive Order No 11,850, 40 Fed. Reg. 16,187 (Apr. 8 1975): 143 Cong. Rec. S3657.

[60] Dolnik, A and Pilch, R (2003) The Moscow theater hostage crisis: The perpetrators, their tactics, and the Russian response. International Negotiation, Vol 8, p 581.

[61] Ibid, p 582.

[62] ICRC (2012) Report of an Expert Meeting. “Incapacitating chemical agents”: Law enforcement, human rights law and policy perspectives, Montreux, Switzerland, 24-26 April 2012, p 47-48,

[63] Wax, P, Becker, C, and Curry, S (2003) Unexpected “gas” casualties in Moscow: A medical toxicology perspective. Annals of Emergency Medicine, Vol 41, No 5, May 2003, p 700.

[64] The European Court of Human Rights, Finogenov and others v. Russia, 04/06 2012, 20 December 2011, Conclusions. See also ICRC (2012) Report of an Expert Meeting. “Incapacitating chemical agents”: Law enforcement, human rights law and policy perspectives, Montreux, Switzerland, 24-26 April 2012, p 48,

[65] CWC Art. II.7. See also Michael Crowely, Chemical Control: Regulation of Incapacitating Chemical Agent Weapons, Riot Control Agents and their Means of delivery, Palgrave Macmillan, Basingstoke, 2016, Chapters 6- 11.

[66] Krutzsch, W and Trapp, R (2014) Article II: Definitions and Criteria, in: Krutzsch, W, Myjer, E and Trapp, R (eds.) The Chemical Weapons Convention: A Commentary Oxford University Press, p 89, see also footnote 45.

[67]  ICRC (2012), p 15.

[68] See more on ICA in f.ex.

[69] Art II.2, CWC.

[70] See Michael Crowely, Chemical Control: Regulation of Incapacitating Chemical Agent Weapons, Riot Control Agents and their Means of delivery, Palgrave Macmillan, Basingstoke, 2016, Chapter 5.

[71] Neill, D (2007), p 12.

[72] Ibid.

[73] Krutzsch, W and Trapp, R (2014), pp 71-72.

[74] ICRC (2012), pp 57-58. See also Michael Crowely, Chemical Control: Regulation of Incapacitating Chemical Agent Weapons, Riot Control Agents and their Means of delivery, Palgrave Macmillan, Basingstoke, 2016, Chapters 4 and 5.

[75] Deprivation of life shall not be regarded as in icted in contravention of this Article when it results from the use of force which is no more than absolutely necessary:

(a)  in defence of any person from unlawful violence;

  1. (b)  in order to effect a lawful arrest or to prevent the escape of a person lawfully detained;
  2. (c)  in action lawfully taken for the purpose of quelling a riot or insurrection.

[76] The European Court of Human Rights, Finogenov and others v. Russia, 04/06 2012, 20 December 2011, para 266.

[77] Ibid, p 47.

[78] The European Court of Human Rights, Finogenov and others v. Russia, 04/06 2012, 20 December 2011, Conclusions, para 3.

[79] Ibid, p 53.

[80] Ibid, p 48. The Court rejected the claim that the concentration of the gas had been miscalculated. The judgment notes that, whie the dose used was insufficient to put everyone to sleep, it somehow managed to kill 126 persons.

[81] Neill, D (2007), p 12.

[82] ICRC (2012), p 43. Recall also that the case was decided solely on the basis of Convention for the Protection of Human Rights and Fundamental Freedoms, not on the basis of the CWC.

[83] See the CWC Article I (1) (b). See also Krutzsch, W and Trapp, R (2014), p 70.

[84] Prosecutor v. Duško Tadić a/k/a “Dule”, Decision on the Defence Motion for Interlocutory Appeal on Jrusdiction, 2 October 1995, para. 70.

[85] See Prosecutor v. Ljube Boškoski and Johan Tarčulovski, Case No. IT-04-82-T, Judgement, 10 July 2008, para. 177; Prosecutor v. Vladimir Ðorđević, Case No. IT-04-87/1-T, Judgement, 23 February 2011, para. 1523.

[86] See Ðorđević Trial Judgement, para. 1526.

[87] ICRC Commentary Geneva Convention I, para. 426 et seq.

[88] Tadić Jurisdiction Decision, para. 70.

[89] ICRC Commentary Geneva Convention I, para. 486, footnote 218. The authorities cited include: ICTY, Haradinaj Trial Judgment, 2008, para. 100; ICTR, Akayesu Trial Judgment, 1998, para. 619; Rutaganda Trial Judgment, 1999, para. 92; and ICC, Lubanga Trial Judgment, 2012, paras 533 and 548. It has also been reflected in State practice; see e.g. United Kingdom, Manual of the Law of Armed Conflict, 2004, para. 15.3.1; Council of the European Union, Independent International Fact-Finding Mission on the Conflict in Georgia, Report, Vol. II, 2009, pp. 299–300; Colombia, Constitutional Court, Constitutional Case No. C-291-07, Judgment, 2007, para. 1.2.1. For a discussion of the notion of ‘peaceful settlement’, see e.g. Jean-François Quéguiner, ‘Dix ans après la création du Tribunal pénal international pour l’ex-Yougoslavie: évaluation de l’apport de sa jurisprudence au droit international humanitaire’, Revue internationale de la Croix-Rouge, Vol. 85, No. 850, June 2003, pp. 271–311, at 282–283.

[90] ICRC Commentary Geneva Convention I, paras. 489-491.

[91] Ibid., para. 495.

[92] David Johnson, “Separatists Kill 18 Pro-Russians in Chechnya, Putin Says War Is Over”, Russia Weekly, 19 April 2002,; “Chechen Peace Amid Gunfire”, CNN, 21 December 2002,

[93] Bhattacharji, P (2010) Chechen terrorism (Russia, Chechnya, separatist).  Council on Foreign Relations, 8 April, 2010,

[94] Tadić Jurisdiction Decision, para. 70.

[95] ICRC Commentary Geneva Convention I, para. 459, footnote 172. The authorities cited include: See e.g. Germany, Federal Prosecutor General at the Federal Court of Justice, Fuel Tankers case, Decision to Terminate Proceedings, 2010, p. 36: ‘[T]he the basic objectives of international humanitarian law and the practical impossibility of differentiating in this context tend to support the conclusion that, in principle, a subject of international law such as Afghanistan – including its allies – cannot be involved in a non-international armed conflict except as a territorial whole.’ See further David, pp. 261–262, and Kleffner, 2013b, p. 59. For the purposes of humanitarian law, a State’s territory includes not only its land surface but also rivers and landlocked lakes, the territorial sea, and the national airspace above this territory.

[96] ICRC Commentary Geneva Convention I, para. 460.

[97] Ibid., para. 461.

[98] Ibid., para. 462.