Arbitrary Detention in Non-International Armed Conflicts: A Tale of Two Hague Courts

Written by and

On 3 April 2023, the much-anticipated trial against Hashim Thaçi, Kadri Veseli, Rexhep Selimi and Jakup Krasniqi started at the Kosovo Specialist Chamber (‘KSC’). In its opening statement, the Prosecution told the judges that each accused is criminally responsible of four counts of war crimes – to wit: illegal and arbitrary arrest and detention, cruel treatment, torture and murder – and six counts of crimes against humanity including imprisonment, torture, murder, persecution and other inhumane acts. In particular, the Prosecution alleges that the four accused, who were all founding members of the Kosovo Liberation Army (‘KLA’), participated in a joint criminal enterprise (‘JCE’), the common objective of which was to ‘gain and exercise control over all of Kosovo by means including unlawfully intimidating, mistreating, committing violence against, and removing those deemed to be opponents.’ (Indictment, para. 32) This JCE was in existence between March 1998 – September 1999 and, according to the Prosecution, either ‘encompassed’ the abovesaid crimes or, in the alternative, some of them were a foreseeable consequence of its execution. (Ibid., paras 32 and 34).

Those who followed the pre-trial proceedings in this case would know that the Defence and the Prosecution have already clashed on a host of interesting legal questions that were raised as jurisdictional challenges (see here and here) and are likely to resurface in the trial proceedings ahead. One of them is the question of whether arbitrary detention in a non-international armed conflict (Count 2 of the Indictment) was a war crime under customary international law during the 1998-1999 armed conflict in Kosovo. This contribution examines the novel line of reasoning that KSC judges have used so far to answer this question and compares it to that used in a recent case from the Netherlands: i.e., the trial of Abdulrazaq R. (Abdulrazaq Judgment) In that case, a local Dutch court applied extraterritorial (universal) jurisdiction to convict an Afghan national of this war crime, committed during the 1980s civil war in Afghanistan. For reasons of brevity, we do not discuss here the interesting definitional challenges that arise when defining the legal elements of ‘arbitrary detention’ in a NIAC. Rather, our focus falls exclusively on studying the methodology which these two courts adopted to establish that arbitrarily depriving civilians of their liberty in NIACs is, by itself, a war crime under customary international law.

Charging arbitrary detention in NIACs as a war crime: the story at the KSC so far

The crime listed under Count 2 of the Thaçi et al. Indictment has already prompted quite some jurisprudential traffic at the KSC, with several accused arguing that no such war crime existed under customary international law during the Kosovo War. (here, paras 131-50; here, paras 46-60). It bears noting here that none of the modern international criminal courts and tribunals has ever found that arbitrarily depriving a civilian of their liberty in NIACs is, by itself, a war crime. To be sure, when committed in an IAC, this conduct could constitute ‘unlawful confinement’, as a grave breach of the Geneva Conventions (e.g., Delalić et al. Appeal Judgment, paras 315-322). Also, when committed in a widespread or systematic attack against a civilian population, such conduct may further qualify as the crime against humanity of ‘imprisonment’ (e.g., Kordić & Čerkez Appeal Judgment, para. 69). However, no international treaty or statute – and not the KSC Law – expressly establishes that ‘arbitrary detention’ is a war crime in NIACs. Some have viewed this as a potential gap in the protection granted to civilians in NIACs, compared to IACs (Dingwall, at 139-40). More generally, IHL experts have long pointed out that one of the least regulated areas of IHL are exactly the rules governing security detention of civilians in NIACs. A recent seminal study on this topic notes that IHL’s ‘silence on the legality of detention [in a NIAC] leaves a normative vacuum that non-State armed groups logically need to fill with their own understanding of regulatory framework for internment.’ (Plamenac, at 3)

The first judicial analysis on whether the crime listed under Count 2 of the Thaçi et al. Indictment amounted to a war crime under customary international law during the Kosovo War was provided by Judge Guillou in a Decision on Motions Challenging the KSC Jurisdiction that he delivered on 22 July 2021. He first found that Article 14(1)(c) KSC Law implicitly provides a legal basis for charging arbitrary detention in NIACs as a war crime. This provision, which is nearly identical to Article 8(2)(c) Rome Statute, establishes jurisdiction over serious violations of Common Article 3 (‘CA3’) of the Geneva Conventions. The one important difference in their texts is that, before listing the otherwise well-known ‘serious violations’ under CA3, the Rome Statute provision states ‘namely, any of the following acts’, while the KSC Law provision states ‘including any of the following acts.’ According to Judge Guillou, this language indicates that Article 14(1)(c) KSC Law offers a non-exhaustive list of ‘serious violations’ of CA3, of which ‘arbitrary detention’ in a NIAC can be one. (Decision, paras 143-146)

He then went on to answer positively the question of whether arbitrarily detaining civilians in a NIAC constituted a war crime under customary IHL during the indicted times. The crux of his analysis was that: (i) CA3 establishes a requirement to treat humanely all persons not taking active part in the hostilities, and (ii) detaining such a person arbitrarily – i.e., without legal basis to do so, or without providing basic procedural safeguards – violates that core IHL principle of humane treatment, be it in an IAC, or in a NIAC. (Decision, paras 152-153) Judge Guillou cited here Rule 99 of the ICRC’s Study on Customary IHL, which states that arbitrary deprivation of liberty is prohibited in both IACs and NIACs, and ‘is not compatible with this requirement [of humane treatment]’ enshrined in CA3. On this basis, Judge Guillou found that the said conduct constitutes a ‘serious violation’ of CA3, albeit not expressly listed therein. (Decision, para. 156) Being thus ‘read in’ as a serious violation of CA3, the said conduct entails individual criminal responsibility and is a war crime. Still, it bears noting that Judge Guillou also referred to nearly 20 states that (ostensibly) criminalized ‘arbitrary detention’ in NIACs, citing the ICRC’s Study, as well as to resolutions by international organizations that condemned this conduct. (Decision, paras 159-166)

Judge Guillou’s analysis on the legal basis for the war crime charged in Count 2 of the Thaçi et al. Indictment was endorsed by the KSC Appeals Chamber in a Jurisdictional Decision from 23 December 2021. Summarizing approvingly the gist of his reasoning, it held:

The Panel finds that […] it was established at the time of the alleged crimes that arbitrary detention was considered to be a serious violation of the principle of humane treatment. Since the requirement of humane treatment is enshrined in Common Article 3 and arbitrary detention is incompatible with this principle, it follows that arbitrary detention entails criminal responsibility. The Panel therefore finds that Veseli has failed to show any error in the Pre-Trial Judge’s finding that arbitrary detention was criminally prohibited under CIL in the context of a non-international armed conflict. (Ibid., para. 109)

Emerging Dutch jurisprudence on the war crime of arbitrary detention in NIACs

Around the same time KSC judges were breaking new ground in war crimes law by finding that arbitrary detention in NIACs is a serious violation of CA3, a domestic Dutch court situated just 2 kilometers away from them was examining the same question. This was the District Court of the Hague (‘the Court’), trying an Afghan asylum seeker, Abdulrazaq R., on war crimes charges covering events that took place during the 1980s civil war in Afghanistan. The Court delivered its judgment on 14 April 2022. It found that between 1983 and 1988, the accused held a senior managerial position in the Pul-e-Charkhi prison, in which the Afghan State Intelligence Agency (‘KhAD’) detained political prisoners. They were held there in deplorable conditions for years, either without trial, or having received minutes-long trials without procedural safeguards. (Ibid., Sections 12-13)

Under Dutch criminal law, it is not possible to charge crimes against humanity for acts committed before the Dutch International Crimes Act (‘Wet Internationale Misdrijven’) entered in force in 2003. Abdulrazaq R. was thus only charged with war crimes, one of which ‘arbitrary detention’, which the Prosecution formally based on the provisions of Common Article 3 of the GCs and on customary international law. (Abdulrazaq Judgment, Section 2) It also bears noting that the indictment also had separate war crimes charges of ‘cruel treatment’ and ‘outrages upon personal dignity’, as a serious violation of Common Article 3, which focused on the mental and physical suffering that the prisoners sustained as a result of the deplorable prison conditions in Pul-e-Charkhi. The purpose of the ‘arbitrary detention’ war crimes charge was thus to establish that arbitrarily depriving a civilian of their liberty in a NIAC is per se a war crime.

The Court’s findings on the legal basis for this war crime charge are largely identical to the aforesaid line of reasoning adopted by the KSC. The Dutch judges held that even though it is not expressly listed in CA3, this conduct is contrary to the principle of humane treatment and therefore falls within the ambit of this provision. Once again, the bread and butter of the analysis was Rule 99 of the ICRC’s Customary IHL Study (here, at 344-253), which the Court cited to conclude that ‘the prohibition on arbitrary deprivation of liberty in the period referred to in the indictment belonged to customary international law and is contrary to the injunction on humane treatment [in CA3].’ (Judgment, Section 17) Reference was also made to the ICRC’s findings that about 70 states criminalize arbitrary detention in armed conflicts, many of which (arguably, see infra) also in the context of NIACs. (here, at 347) Overall, the Court’s analysis on this issue is less than a page long and resulted in the conclusion that ‘arbitrary detention’ was a war crime also in NIACs under customary international law during the 1980s.

A call for caution: understanding the significance of the KSC and Hague District Court’s findings

While no international court or tribunal has ever held that ‘arbitrary detention’ in a NIAC is, by itself, a war crime, this conduct was subject to war crimes litigation at the UN Tribunals. Indeed, it was part of the war crimes charges against three KLA members in the ICTY Limaj et al. case. Significantly, however, was legally qualified as the war crime of ‘cruel treatment’, which is one of the serious violations listed in CA3 (Limaj et al. Trial Judgment, para 232). This war crime has its own constituent elements, which require inter alia an act or omission that ‘causes serious mental or physical suffering or injury or constitutes a serious attack on human dignity.’ (Blaškić Appeal Judgment, para. 595). In Limaj, the ICTY judges held that:

In determining whether the “unlawful seizure”, “unlawful detention for prolonged periods” and “interrogation” alleged in the instant case amount to cruel treatment, the Chamber has, therefore, taken into account all the circumstances of the instant case. The Chamber has come to the conclusion that, at least in the circumstances of this case, these acts in and of themselves do not amount to a serious attack on human dignity within the meaning of cruel treatment under Article 3 of this Statute. Count 2 must therefore also be dismissed. (Limaj et al. Trial Judgment, para 90).

Thus, while they did not exclude the possibility that in some circumstances the act of unlawfully detaining a person in a NIAC may satisfy the elements of ‘cruel treatment’, the judges evidently rejected the notion that this act automatically constitutes an attack on human dignity that would constitute the war crime of ‘cruel treatment’ under CA3. The appalling conditions in which the KLA (arbitrarily) detained the victims were used instead to establish the commission of the said war crime. (Ibid., paras 288-289)

Recognizing ‘arbitrary detention’ as a separate and distinct war crime in a NIAC clearly challenges that reasoning and seeks to affirm that the very act of arbitrarily depriving a protected person of their liberty is, by itself, grave enough to qualify it as a war crime. The methodology that the KSC and the Hague District Court used to confirm this, however, raises some concerns.

First, no international court or tribunal has ever accepted that CA3 is an open-ended provision, which criminalizes any unlisted conduct that is deemed to breach the general IHL ‘principle of humane treatment’. It is certainly true, as the KSC Appeals Chamber stressed, that the Tribunals have previously considered various types of unlisted conduct (e.g. ‘forced labour’, ‘maintaining an atmosphere of terror’, ‘abduction of women and girls as bush wives’, etc.) to fall in the scope of CA3. (Jurisdictional Decision, para 96, fn 267) However, in all such instances, the approach was always to determine whether the identified conduct satisfied the specific legal elements of one of the expressly listed war crimes (‘serious violations’) under CA3: be it ‘cruel treatment’, or ‘outrages upon personal dignity’. Whenever the UN Tribunals sought to establish a separate war crime that is not listed in their statutes – like e.g., the ICTY Galić Trial Chamber did with the war crime of terrorizing the civilian population (Galić Trial Judgment, paras 89-138) – they always did so by systematically applying the well-known four ‘Tadić conditions’, namely:

    1. the violation must constitute an infringement of a rule of international humanitarian law;
    2. the rule must be customary in nature or, if it belongs to treaty law, the required conditions must be met;
    3. the violation must be “serious”, that is to say, it must constitute a breach of a rule protecting important values, and the breach must involve grave consequences for the victim; and
    4. the violation of the rule must entail, under customary or conventional law, the individual criminal responsibility of the person breaching the rule (Tadic Interlocutory Appeal, 94).

Second, reading new war crimes in CA3 by stating that a given unlisted conduct violates the general IHL ‘principle of humane treatment’ captured in this provision, and thus also entails individual criminal responsibility, risks offending the principle of legality. Indeed, as the ICRC also notes, the principle of humane treatment is ‘an overarching concept’, for which ‘[t]he actual meaning […] is not spelled out, although some texts refer to respect for the “dignity” of a person or the prohibition of “ill-treatment” in this context.” (Rule 87 of ICRC Customary IHL Study). A whole lot of IHL norms that are generally meant to ensure respect for the dignity of protected persons – yet do not necessarily involve grave consequences for the victim, nor entail individual criminal responsibility under custom or treaty – may be re-interpreted as war crimes though the short route described in paragraph 109 of the KSC Appeals Chamber’s Jurisdictional Decision.

Finally, it goes without saying how invaluable the ICRC’s Study on Customary IHL is, as well as how authoritative it is. For domestic courts trying war crimes, in particular, it is often a lifeline since they standardly lack the international staffing and resources necessary to conduct independent and thorough studies of state practice and opinio juris required by e.g., the ‘Tadić conditions’-analysis. At the same time, however, it is important to keep in mind that the ICRC is not a judicial organ and the findings in its Study of Customary IHL ought to be independently verified by courts of law when deciding on the innocence or guilt of an accused. For instance, one state which is listed by the ICRC as having legislation that proscribes arbitrary deprivation of liberty as a war crime in NIAC, and which was also referred to by the KSC Appeals Chamber, is Bulgaria. (ICRC Customary IHL Study, at 347, fn 290; Jurisdictional Decision, para 106, fn 297) The ICRC cites to this end Article 412 of the 1968 Bulgarian Penal Code which, however, implements the grave breaches from Article 147 of GCIV within the Bulgarian legal order. The chapeau of the said Article 412 starts with the sentence ‘[a] person who in violation of the rules of international law for waging war’ and then lists certain acts that are criminal when committed against civilians. It is quite unlikely that the Bulgarian government pursued a progressive stance on the development of IHL by extending the meaning of the word ‘war’ (‘война’) in this article beyond the international armed conflicts to which its Geneva Conventions counterpart applies.

One last point should be stressed here: we are not arguing that the two Courts are wrong to conclude that ‘arbitrary detention’ in a NIAC was a serious violation of the laws and customs of war during the indicted periods. Although the judges did not structure and conduct their legal analysis along the lines of the Tadić conditions, it would be possible to piece together from their findings much of what would otherwise be necessary to conclude that these criteria are met by the said conduct. It has also already been argued in academia that arbitrarily depriving civilians of their liberty in NIACs satisfies these ‘Tadić conditions’ for being recognized as a war crime. (Ventura, at 155-57) Our point is, rather, that the methodology that these two Courts adopted creates a dangerous precedent.


The development of international criminal law (‘ICL’) has often been described as a ‘laboratory process’, in which the international courts and tribunals have sometimes had to experiment with and recalibrate the definitions of substantive and procedural ICL norms. (Whiting; Christensen and Orina). This is quite understandable, given how young this field of law still is, yet it is also important that the said process is properly balanced against the dictates of the legality principle. Whenever that balance has not been met, strong criticism of judicial law-making has ensued.

For the purpose of declaring ‘new’ (i.e., not listed in statutes, but part of customary ICL) war crimes, the four conditions that the ICTY Tadić Appeals Chamber first specified form the analytical roadmap that should be followed in order to maintain the above-said balance. In this sense, the present authors are of the view that the methodology used by the KSC and the District Court of the Hague to establish ‘arbitrary detention’ in NIAC as a war crime – namely, reading Common Article 3 of GCs as an open-ended provision, which allows unspecified conduct to be by itself considered a war crime if it is deemed incompatible with the ‘principle of humanity’ – is problematic. At the same time, when considering specifically the work of national courts that are trying core international crimes, the question naturally arises of how realistic and feasible it is that a full-blown ‘Tadic conditions’-analysis (specifically, on elements (ii) and (iv)) could be carried out by a chamber of e.g., Dutch judges who have many other (non-ICL) cases to process. Relying on the ICRC’s seminal study of customary IHL is highly recommended, so long as this does not translate into automatic, uncritical acceptance of the ICRC’s findings. These are some of the challenges that domestic courts, in particular, would need to wrestle with when taking on the task of developing ICL beyond the parameters established by the ICC and the UN Tribunals.

Print Friendly, PDF & Email

Leave a Comment

Comments for this post are closed


Michael G. Karnavas says

April 18, 2023

Excellent analysis. Exceptionally measured and informative.

Lachezar Yanev says

April 19, 2023

Many thanks, Michael

DH says

April 20, 2023

Thank you very much for this very interesting and thought provoking piece.
For me, it is fascinating that both Court concluded that 'arbitrary detention' in NIAC may amount to a war crime, and by that, arguably, recognizing that there exists a legal basis (or at the very least, power) to detain in NIAC. As the authors of this piece may be aware, one of the many controversies in contemporary IHL is whether IHL, and in particular the law of NIAC, provide for a legal basis to detain civilians for security needs. Had the KSC and the Dutch District Court followed the proposition that there is no legal basis to detain in NIAC, and therefore any detention, is ipso facto, an arbitary deprivation of liberty, they would not delve into an analysis on the matter. So, one could argue, that these rulings strenghthen those who submit that IHL (CA3) provides an implicit legal basis to detain in NIAC. In other words, if some form of detention in NIAC may be arbitrary, so as to constitute a war crime, then if also follows that there must be detention which is not arbitrary in NIAC.
Again, many thanks for this interesting piece.

Asher Rottenberg says

April 24, 2023

I am not so sure any of the judgments can shed additional light of the question of legality of detention of civilian in NIAC under IHL. Even if any civilian detention is prohibited under IHL – it does not mean that it constitutes a criminal offence amounting to War Crime. Following the courts' rationale, as presented in the excellent blogpost, if one bases herself of the inhumane character enshrined in CA3, then only arbitrary detention which can be linked with inhumane treatment should be classified so for criminal purposes. I can easily see how one develops an argument (although I do not have the time to do it myself!) that only certain detentions are arbitrary for the purpose of criminally, but others can be prohibited under IHL and still not amounting to inhumane treatment that qualifies them as a criminal offence.
This is also true with regard to the other argument which is based on state practice (imported via the ICRC study). One has to look into these statues and surely arbitrary detentions enshrined in domestic legislations refers to arbitrary detention without any domestic basis, and not one with a domestic basis (which does not necessarily affect their legality under international law).