The legal and practical issues related to the exceptional practice of internment of protected persons under the Fourth Geneva Convention are complex. One such question is when internment begins. The treaty provides guidance on grounds for internment and the procedural safeguards to be applied, as well on the requisite conditions of internment, but is silent on when this type of detention actually starts. This gap in the law, which is the focus of examination, has proven time and again to have pernicious effects on the protection of detained civilians. 

In this post, and as part of a series on the rules governing the grounds and procedures for the internment of protected persons, former ICRC Senior Legal Adviser Jelena Pejic suggests that a detained civilian should be deemed an internee no later than two weeks after being deprived of liberty for reasons related to an armed conflict, if not released earlier or designated a criminal suspect.

Editor’s note: The forthcoming update of the Commentary on the Fourth Geneva Convention – an ICRC project seeking to provide current interpretations of the Conventions to enhance their understanding, dissemination and faithful application in today’s armed conflicts – will consider and address the rules explored in this blog series.

The conduct of hostilities and the detention of persons fighting for the adversary are the main ways in which a party to an armed conflict may seek to weaken and ultimately prevail over an enemy. Internment is a type of detention specific to armed conflict which, like the broader notion of detention, is not defined in international humanitarian law (IHL).[1] It is generally understood to mean the non-criminal detention of a person based on the serious security threat their activity poses to the detaining authority.

A large part of the Third Geneva Convention applicable in international armed conflict (IAC) deals with the internment of prisoners of war (POWs) i.e., of members of the armed forces of a party to such a conflict who have fallen into the power of the enemy whether by capture, surrender or capitulation. The internment of POWs begins immediately upon their falling into enemy hands and may last until the end of active hostilities, after which they must be released and repatriated. No process for assessing the lawfulness of POW internment is provided for in the Third Geneva Convention, because POWs are for the most part captured combatants and are recognizable because of their uniform. They are not only likely to go back to the fight if released, but in many cases have a duty under domestic law to attempt to escape and rejoin their forces even if interned. In other words, POWs are inherently a security threat.

The legal and practical issues related to civilian internment under the Fourth Geneva Convention (GC IV), also applicable in IAC, are far more complex. One such question is when internment begins. The treaty provides guidance on the grounds for internment and the procedural safeguards to be applied, as well as on the requisite conditions of internment (the latter of which will not be further addressed below), but is silent on when this type of detention actually starts. This gap in the law, which is the focus of examination here, has proven time and again to have pernicious effects on the protection of detained civilians.

By way of reminder, GC IV specifies that internment of civilians may take place in a state’s own territory “only if the security of the Detaining Power makes it absolutely necessary” (GC IV, Article 42, para. 1). For occupied territories, the treaty provides that “If the Occupying Power considers it necessary, for imperative reasons of security, to take safety measures concerning protected persons, it may, at the most, subject them to […] internment” (GC IV, Article 78, para. 1).  Leaving aside discussions of whether there is any difference between the grounds of absolute necessity and imperative reasons of security, it is clear that, in both scenarios, the legal standard is high.

GC IV does not state what specific conduct by a civilian may be deemed to reach the requisite level of security threat, and no list of acts could comprehensively reflect the various ways in which a Party’s safety may be seriously undermined in practice. There is no doubt that a civilian’s direct participation in hostilities would meet the legal threshold,[2] but other acts may do so as well (general recruitment, intelligence-gathering, or financing of military operations by civilians, etc.). The onus is on the detaining state to properly determine what particular conduct meets the bar for internment. This should be done in good faith, also keeping in mind that unnecessary detention is not operationally useful, drains resources, saps staff morale and can affect perceptions about the legitimacy of a military campaign as a whole.

The Fourth Convention also provides rules on the procedural safeguards that must be applied, in each individual case, to determine whether a detained civilian represents a security threat warranting internment. The formulation of the relevant rules differs for a state’s own territory, and occupied territory. The broad gist of both is that after an “order” or “decision” on internment is taken by the Detaining/Occupying Power, a detainee has the right to have such determination promptly reviewed by a competent body – an administrative board or court – and thereafter every six months. The purpose of the initial, and subsequent, review(s) is to establish whether the specific reasons which necessitated internment still exist. If that is not the case, the internee must be released. Civilian internment must always end after the close of military operations.

There is nothing in the Convention to suggest that internment begins from the moment a protected person falls into the hands of the enemy.  In reality, there are different scenarios in which civilians may be deprived of liberty, such as in a combat zone, in a house search, or at a check point. The information based on which they are deemed to pose a threat may be directly observed or may be based on various types of intelligence, whether genuine or, as has been known to happen, supplied by malicious informants from the same community. Civilians may be taken from their homes by the opposing side for nothing more than the work or functions they perform, or swept up and detained after security incidents against enemy forces near their homes. The upshot is that, as opposed to members of the armed forces who are considered to be interned upon falling into enemy hands based on their very status, there are myriad reasons for which civilians may be deprived of liberty, meriting further examination.

Determining when internment begins in the case of civilians is necessary in order to prevent their possible arbitrary detention, or to use the relevant IHL term, their unlawful confinement, which is a grave breach of GC IV  and therefore prosecutable by States Parties under various forms of criminal jurisdiction, including universal. Importantly, unlawfulness in this context does not refer to possible violations of IHL rules on treatment of internees in the “narrow” sense (e.g. murder, torture, medical experiments, etc.), but to non-implementation of the grounds and process for internment as specified in the Fourth Convention.

What is at issue here is that unless and until an internment order/decision is taken and the procedural rules on internment “kick in”, civilians may be detained in a legal limbo, not necessarily knowing why they are being held or what procedural safeguards they are due. Regrettably, the reality on the ground does not, for the most part, comply with the relevant legal norms. Detaining authorities may simply not issue internment orders and set up an internment review process, but hold civilians on alleged security reasons for unspecified periods of time with no review, and, ultimately, no possibility of release before the end of the conflict, as foreseen by GC IV.

Given the lack of guidance in IHL on when an internment order or decision must be issued, it would be optimal if IHL were interpreted to mean that any civilian detained is to be immediately considered an internee, or a criminal suspect if the requisite conditions for the latter type of detention have been fulfilled. That, however, is neither the law, nor the practice in IAC. This is because the activities of a civilian believed to pose a security threat (or to have committed a criminal act) may in practice take time to piece together and properly assess and may include the need to examine their joint conduct with other persons.

It is nevertheless submitted that internment should be assumed to have begun at an identifiable point in time in order to prevent the procedural legal limbo mentioned above. Based on a timeframe found in Article 136 of the Fourth Convention, it is suggested that a detained civilian should be deemed an internee no later than two weeks after being deprived of liberty for security reasons related to an armed conflict, if not released earlier or designated a criminal suspect.

Pursuant to Article 136,  the parties to an IAC must establish an official Information Bureau for receiving and transmitting information related to protected persons who are in its power. In addition: “Each of the Parties to the conflict shall, within the shortest possible period, give its Bureau information of any measure taken by it concerning any protected persons who are kept in custody for more than two weeks, who are subjected to assigned residence or who are interned” [emphasis added].

Due to somewhat unclear drafting, this part of the article is open to interpretation. It does not specify whether “persons kept in custody for more than two weeks” is a category separate to those interned or subjected to assigned residence (in which case an additional ‘or’ should have been inserted), or whether the latter are illustrations of the former. The ICRC’s 1958 Pictet Commentary suggests the former. Based on the proceedings of the Conference of Government experts summarized by him, it means that information pertaining to civilians detained on a criminal offence, as well as those interned or subjected to assigned residence must be communicated by each state to its National Information Bureau. As explained in the Commentary: “From the beginning of its studies, the International Committee of the Red Cross had considered that the general regulations governing internment should also be applied to civilians arrested either for offences against the Occupying Power or for security reasons, and that their names and the fact of their arrest should be communicated to the Power of origin. […] The keeping of such a record, however, would not be necessary if detention did not exceed two weeks”.

The ensuing text of Article 136 thus allows for a period of time to elapse before notification to the National Information Bureau of a civilian deprived of liberty becomes obligatory, based, inter alia, on the need of a detaining power to determine the type of activity in which such person may have been engaged.[3] However, the wording of the Article implies that an internment order or decision will have been taken within two weeks. It is submitted that if such an order or decision is not handed down in that time-frame, it should be assumed that internment has begun at the expiry of that period. As already mentioned, a cut-off moment is necessary for reasons of procedural protection. It ensures that a detained civilian is able to benefit from the internment process provided for in GC IV, and that a party to an IAC cannot circumvent the relevant rules simply by failing to classify a deprivation of liberty as internment.

A final, anticipatory remark may be made. Resistance to the approach suggested here may be expected from some military practitioners based on arguments about the extra resources, logistics and time that the proper implementation of an internment framework would require. This may be true. But the procedural safeguards at issue are legal obligations. Moreover, IAC is in any case a costly endeavor. It is difficult to accept that the necessary resources, logistics and time can be found for other aspects of military operations, but not for the protection of civilian detainees whose position is inherently vulnerable. It is, simply put, a question of humanity.

[1] For the purpose of this discussion, detention is understood to mean the holding of a person in a bounded space which they are not free to leave at will.

[2] Under IHL, direct participation in hostilities may result in a civilian being lawfully targeted and killed, or may lead, inter alia, to internment. It is important to properly determine what conduct constitutes direct participation, as any and all conduct that does not meet that description cannot lawfully entail the taking of life, but can be dealt with through detention if the requisite conditions are met.

[3] See, e.g, Pictet Commentary to GC IV, Article 5: “Article 136 lays down, for example, that the names of the detained persons are to be transmitted if they are kept in custody for more than two weeks: one can see that this leaves a margin which will, in the majority of cases, meet any legitimate security requirements”.

See also:

• Mikhail Orkin, The internment of protected persons and the Fourth Geneva Convention, May 16, 2024
• Giulio Bartolini, Voluntary reports: a new tool ‘toward a universal culture of compliance with IHL’, April 11, 2024
• Jacob Coffelt, Codifying IHL before Lieber and Dunant: the 1820 treaty for the regularization of war, April 4, 2024
• Tilman Rodenhäuser, Samit D’Cunha, Foghorns of war: IHL and information operations during armed conflict, October 12, 2023