Torture in Libya and Questions of EU Member State Complicity

Amnesty International has reported that ‘tens of thousands’ of refugees and migrants are being subject to torture and other human rights abuses at the hands of Libyan state officials and non-state actors operating in, and out of, Libya (the full report can be accessed here). The publication of the report has led to allegations that the European Union (EU) is complicit in torture. One finding of the report is that ‘EU member states are and have been well aware of the widespread human rights violations and abuses suffered by refugees and migrants in Libya’ (p. 56). Amnesty International has claimed that EU states ‘are complicit’ in torture. Whether the complicity spoken of can trigger the responsibility of these states under international law is implied, but far from clear.

There are many tangents to questions of ‘European complicity’ in the torture of Libyan refugees and migrants. For example, issues regarding the obligation of non-refoulement (p. 53 of report), or the extraterritorial application of human rights obligations (pp. 54-56) (for insights on these particular matters see Gauci and Jackson respectively). The following post will briefly analyse the applicable secondary rules relating to how EU states could be held responsible for complicity in torture under general international law in light of the facts contained in the Amnesty report.

The relevant legal framework

Article 2 of the International Law Commission (ILC) Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA) clarifies that state responsibility can be triggered if an act or omission is attributable to a state and also breaches an international obligation of that state. Article 16 of the ARSIWA sets out two conditions for holding states responsible for aiding or assisting another state in an internationally wrongful act:

(a) The state ‘aiding or assisting’ another state must do so with knowledge of the wrongful act in question.

(b) The act would be internationally wrongful if committed by that state.

How these requisites of state responsibility for complicity can be met in this particular case pose a number of issues. For the purposes of this post, the ‘aid or assistance’ spoken of is the funding, technical and logistical support given to Libyan border control authorities by EU states. As the Amnesty report provides substantial evidence in this regard (pp. 42-50), it shall be assumed that there is a sufficient nexus between this ‘aid or assistance’ of EU states, and its contribution to the acts of torture in Libya.

The knowledge requirement

The mental element of Article 16 is not as straightforward as it seems at face value. There are in fact three possible ways it could be satisfied:

1. Proving that the EU state knew of the torture in Libya (Knowledge).
2. Proving that the EU state intended to facilitate the torture in Libya (Purpose).
3. Proving that the EU state avoided knowledge of the torture in Libya (Wilful Blindness).

Preference towards one interpretation depends on how narrowly one wishes to construe the mental requirement. If one strictly adheres to the language of Article 16 alone, then option 1 appears to be correct, whilst leaving the door open for a broad application of the provision. However, upon examination of the provision’s commentary (particularly paras. 1 and 5), it becomes clear that the ILC had in mind an intention to facilitate wrongful conduct, thus favouring option 2. This interpretation of Article 16(a) is more restrictive, requiring an additional component to knowledge of the overall circumstances. A number of arguments have been made against such a reading (see Aust p. 236). Yet the provisions of the ARSIWA are best read in conjunction with their commentary to ensure for the most accurate interpretation of the rule under scrutiny. The term ‘wilful blindness’ is absent from Article 16, but has become a point of discussion in issues relating to complicity (see, for example, Jackson, pp. 54 and 162 and Chatham House, pp. 14-17). Applying this concept adds an extra layer of complexity to the knowledge requirement (see point 3).

The three interpretations are interconnected, but the relationship between them will not be analysed further here. Instead, each will be taken separately to show how their application would play out on the facts contained in the Amnesty report.

1. Knowledge

Paragraph 4 of the commentary to Article 16 shows that the knowledge required is that of illegality. This means that the EU states must have been aware of facts proving torture had occurred in Libya, and that such acts were internationally wrongful. The report of Amnesty is comprehensive in its assessment that EU member states are aware of the acts of torture being perpetrated in Libya (pp. 51-59). In particular, it highlights the numerous sources of information compiled by international organisations, non-governmental organisations and states that point towards EU states being aware of the circumstances in which refugees and migrants in Libya are being tortured (pp. 56-58). More specifically, for example, Italy explicitly confirmed its knowledge of such incidents in a letter from its Prime Minister to Amnesty International (p. 58). These sources of information also speak to the clear unlawfulness of torture under international law, meaning no state could make a rational argument that facts showing torture would not necessarily mean that those acts were unlawful. Based on the available facts, EU states are in no position to claim that they are unaware of the present circumstances. They have actual or near-certain knowledge of torture against refugees and migrants in Libya. If the test under Article 16(a) is pure knowledge, then it appears its threshold would be met in this case. What if, however, the test was purpose?

2. Purpose

The commentary to Article 16 indicates that ‘the aid or assistance must be given with a view to facilitating the commission of the wrongful act […] A State is not responsible for aid or assistance under article 16 unless the relevant State organ intended, by the aid or assistance given, to facilitate the occurrence of the wrongful conduct’. The section of the Amnesty report dealing with torture (pp. 31-33) does not indicate EU states are providing aid or assistance with a view to facilitating this wrongful act, and actually doing so. If Article 16(a) requires application of a purpose-based test, the Amnesty report, despite its detail, does not show any facts that would satisfy such a threshold. Adherence to this interpretation thus restricts potential responsibility for complicity considerably in this case. There is simply no evidence that EU states desire migrants in Libya to be tortured. If, however, the purpose test is interpreted as incorporating a more oblique form of intent – for example, that a particular consequence is to be regarded as intended if the relevant state organ is aware that it will occur in the ordinary course of events (cf. Art. 30(2) of the Rome Statute of the ICC) – then the test could potentially be met, since EU states do arguably know that the aid provided will be used for torture in the ordinary course of events in Libya.

3. Wilful Blindness

Wilful blindness in the context of state responsibility for complicity can be defined as a state consciously turning a blind eye to credible information that points to wrongful conduct of another state to which it is aiding or assisting. This test is attached to the knowledge test, but in the negative, in that a state did not have knowledge of wrongfulness, but because of a choice it made to avoid such knowledge. With respect to the torture of refugees and migrants in Libya, a reasoned argument could be made that EU states are turning a blind eye – as suggested by Amnesty (pp. 12, 19 and 58 of report). Applying wilful blindness acts as a potential safeguard to satisfying the knowledge requirement should an EU state claim that it did not know of the torture in Libya. Yet this would be hard to envisage considering the amount of credible information available to EU states and their responses to it.

The opposability requirement and attribution

The second requirement of complicity under state responsibility is opposability: the state that aids or assists the recipient state perpetrating the wrongful act must be bound by the same obligation as that state. To the extent that torture is at issue, the opposability requirement, in the state-to-state context, is straightforward and would be satisfied in this case. Libya is bound by the obligations under the Convention against Torture, as are all EU states. There is also an additional factor that the prohibition of torture is a rule of customary international law of jus cogens status, thereby binding all states. Therefore, on satisfying the mental requirement, EU states could be held responsible for complicity in the acts of torture committed in Libya, subject to one final question: whether it was (1) Libyan organs or (2) non-state actors that tortured the refugees and migrants.

The struggle with 1 is indentifying who are Libyan state officials in reality. The Amnesty report makes clear (pp. 31-33) that the General Directorate for Combating Illegal Migration (DCIM) – part of the Libyan Ministry of Interior – is blameworthy for the instances of torture, which occur predominantly in DCIM detention centres. The status of the DCIM forming part of the Libyan Government makes it logical to presume that the persons working at its detention centres who have tortured refugees and migrants are state organs. Such conduct would therefore be attributable to Libya under Article 4, ARSIWA.

Further issues are presented by 2, which can be addressed through three avenues:

1. Attributing the acts of torture to Libya under one of the rules on attribution of conduct contained in the ARSIWA (Articles 4-11). It should be noted that the conduct of non-state actors could also be attributed to Libya under rules of international law that are lex specialis to the ARSIWA framework (Article 55). If the torture took place at a DCIM detention centre, then non-state actors operating there would likely be acting in complete dependence on Libya (Article 4), or exercising elements of governmental authority (Article 5), or acting on the instructions of, or under the direction or control of, Libya (Article 8). Triggering any of these rules would depend on the facts present in each specific instance of torture. But what about acts of torture that have occurred outside Libyan government installations and cannot be attributed to Libya under one of the rules on attribution of conduct? This is where avenues 2 and 3 become pertinent.
2. Attributing responsibility to Libya through the application of primary rules relating to the positive obligation to prevent torture. The wrongful act here would now not be torture, but the failure to prevent it. Considering the ambiguity surrounding the status of the actors torturing refugees and migrants, and the difficulties of uncovering facts sufficient to satisfy one of the attribution of conduct provisions of the ARSIWA, this avenue is likely to yield the highest chance of success. So long as the torture in question took place within Libyan jurisdiction, it could be argued that the state had failed to fulfil its obligation to prevent torture. Yet what happens if Libya is found not to have breached its positive obligations in this regard – meaning no internationally wrongful act has occurred? There is one last option to pursue.
3. Analogising Article 16 so that it can apply in state-to-non-state interactions. This avenue turns on whether the non-state actor in question bears an international obligation relating to the prohibition of torture. The key issue here is essentially whether there is even the occurrence of an internationally wrongful act for an EU state to be complicit in. EU states have an obligation with respect to the prohibition of torture, but do the non-state actors operating in Libya?

The untapped potential of non-state actors’ obligations

Whether the acts of torture in Libya constitute an internationally wrongful act is a tricky question if they have been committed by a non-state actor. It is unlikely that the militias spoken of in Amnesty’s report are bound by an obligation prohibiting torture. Yet the report also notes the involvement of armed groups in these incidences. It has been established that such non-state actors can, and do, bear obligations under international human rights law, including the prohibition of torture. The qualification for this ability to bear such obligations is whether the non-state armed group is party to an armed conflict. The Libya situation has been described as ‘fluid’, in that it oscillates between an international and a non-international armed conflict. It is likely that the armed groups referred to in the Amnesty report are party to this conflict.

A confirmation that the non-state armed groups are bound by the obligation not to torture advances the argument that EU states could be complicit in an internationally wrongful act of torture if committed by these non-state actors, if the mental requirement were also satisfied. The law in this area is far from fully developed, but exploring the potential obligations of non-state actors in such situations offers the promise of expanding complicity under the law of state responsibility beyond its state-centric confines.

Whether a rule analogous to Article 16 exists in the state-to-non-state context is plausible. The International Court of Justice (ICJ) unlocked the potential for holding a state responsible for aiding or assisting a non-state actor. The Bosnia Genocide judgment (paras. 419-421) set a precedent that the Genocide Convention prohibits complicity in genocide, independent of considerations as to what status the actor committing genocide was under international law. The Court’s ruling was, however, based on the text of the Convention rather than on a generally applicable rule of customary law. Additionally, in its Order for Provisional Measures (p. 25), the ICJ directed Serbia to:

ensure that any military, paramilitary or irregular armed units which may be directed or supported by it, as well as any organizations and persons which may be subject to its control, direction or influence, do not commit any acts of genocide (emphasis added).

There is also state practice to support complicity applying in state-to-non-state interactions. It should be borne in mind that the principal reason for the final version of Article 16 not applying to non-state actors was because it was disputed at the time of drafting whether non-state actors could bear international obligations. Bringing this thought back into the context of the Libyan situation, if this analogy theory were to be accepted, then so long as EU states and non-state actors are bound by the same obligation concerning the prohibition of torture, then state responsibility for complicity in torture by a non-state actor could be invoked. Such an outcome would provide an additional route to triggering state responsibility independent of the ARSIWA framework. So long as the opposability requirement is satisfied, why does the status of the actor, which owes the same obligation as the aiding or assisting state, matter?

Whether EU states can be held responsible for complicity in the acts of torture against refugees and migrants in Libya remains to be seen. There are multiple factors to consider. The Amnesty report makes a strong case that EU states have significantly contributed to circumstances in which people attempting to escape violence are suffering from human rights abuses such as torture at the hands of Libyan state officials and non-state actors. The law of state responsibility has a significant role to play here. This framework of secondary rules has the capacity to ensure that both forms of direct and indirect responsibility are not evaded. The facts surrounding the torture of refugees and migrants in Libya contained in the Amnesty report are substantial and provoke a desire for lawful accountability and justice to ensue. EU states are both morally and politically ‘complicit’ in the torture being inflicted, due to these states arguably turning a blind eye to such acts, or failing to do more to prevent them on a policy level. The legal position is, however, much more complex.