The Ljubljana – The Hague Convention on Mutual Legal Assistance: Was the Gap Closed?

26 May 2023 was a historical day for the fight against impunity for the most serious crimes of concern to the international community. On that day, a Diplomatic Conference of Plenipotentiaries, held in the city of Ljubljana, Slovenia, between 15 and 26 May, adopted by consensus the Ljubljana – The Hague Convention on International Cooperation in the Investigation and Prosecution of the Crime of Genocide, Crimes against Humanity, War Crimes and other International Crimes (“the Convention”). The goal of this treaty is to close a gap in the international legal system: the absence of a multilateral treaty regulating in sufficient detail mutual legal assistance and extradition for the domestic investigation and prosecution of core international crimes (on this gap, see here, here). In addition to eight small annexes, the adopted Convention has 94 articles, divided into eight parts. The treaty was named “Ljubljana – The Hague Convention” in honour of the city in which the negotiations took place (Ljubljana) and the city that will host the official signing ceremony in January 2024 (The Hague).

The Ljubljana Conference was the pinnacle of an endeavour initiated in 2011, when Belgium, the Netherlands, Slovenia, and Argentina launched the so-called Mutual Legal Assistance Initiative (MLA Initiative or Initiative). The project’s goal was to promote discussions and negotiations among states and civil society for the ultimate adoption of that gap-filler treaty. Those four states, later joined by Senegal and Mongolia, established themselves as the “Core Group” of the Initiative, being responsible for spearheading the project.

The MLA Initiative was carried out independently, as a stand-alone process outside the United Nations (UN) or any other organization. This autonomy ensured greater flexibility to the Core Group, but the lack of an established institutional and procedural setting triggered criticism centred on a lack of transparency and inclusivity surrounding the work of the Initiative. In fact, only 68 states (35% of the total UN membership) sent delegations to the Ljubljana Conference and approximately half of them were European.

Between 2018 and 2022, the Core Group circulated five draft conventions following informal consultations between states (here, here, here, here, here). The fifth and most recent text, circulated on 30 November 2022, served as the zero draft for the negotiations in Ljubljana.

This post addresses four relevant elements of the adopted Convention in terms of evaluating its gap-filler purpose: (1) the Convention’s scope; (2) domestic jurisdiction; (3) data protection; and (4) protection of victims. The authors attended the Ljubljana Conference and the observations contained here reflect their personal notes taken during the negotiations.

Scope

The determination of the scope of the Convention was a very divisive issue in consultations prior to the Ljubljana Conference. Several states, spearheaded especially by Switzerland, preferred a “pure MLA” treaty, addressing exclusively extradition and mutual legal assistance as a means to secure wide state adherence to the Convention. Switzerland even circulated its own proposal of draft convention on 12 March 2020.

The Core Group, however, espoused a broader scope to the treaty, aimed at ensuring that the latter constitutes an effective and comprehensive gap-filler instrument. Thus, the Core Group argued that the Convention should cover not only cooperation, but also substantive criminal law issues, particularly the definition of the covered crimes, domestic criminalization, establishment of national jurisdiction, aut dedere aut judicare, and victims’ rights.

Before the COVID-19 pandemic, the pure MLA approach still had some significant support among states. However, in the virtual consultations held in 2021 and 2022, the Core Group saw enough support among states to move forward with the broad scope, as reflected in its zero draft (cf., e.g., Arts. 5, 7, 8, 11-15, 81-83). At the Ljubljana Conference, some delegations, including Greece, Portugal, Australia, Malawi, Switzerland, and the United States (USA), still expressed doubts and concerns about the reach of the Convention. Yet, they did not argue strongly for revisiting the question of scope during the negotiations. It was evident that, by then, the ship had already sailed. Therefore, the question of scope as such was not discussed during the Conference.

In the end, despite intense negotiations in some provisions on non-cooperation issues, such as domestic jurisdiction and victims’ rights, the Conference adopted the Convention by consensus with a broad scope. It remains to be seen if this decision will, in fact, inhibit wide state adherence to the treaty in the future.

Domestic Jurisdiction

The obligation to establish national jurisdiction over the crimes to which the Convention applies was one of the most controversial issues during the Conference. Until late in the evening of 25 May 2023, delegates were still divided over the scope of Article 8 of the Convention. The chances of the treaty being adopted by consensus the following day depended solely on the success of the negotiations regarding this article.

Article 8(1) and (2) were mostly uncontroversial, as they require the states parties to establish jurisdiction on more settled grounds, that is, the place of the crime, nationality of the offender and of the victims and, in case the offender is a stateless person, habitual residence. Article 8(3) was the one that triggered more debate, even though its text is modelled on agreed language from Article 5 of the Convention against Torture (CAT). Article 8(3), as contained in the zero draft, reads as follow:

Each State Party shall likewise take such measures as may be necessary to establish its jurisdiction over such crimes in cases where the alleged offender is present in any territory under its jurisdiction and it does not extradite the alleged offender to any of the States referred to in paragraphs 1 or 1bis, or surrender the alleged offender to a competent international criminal court or tribunal.

This provision addresses the situation where, other grounds of jurisdiction not being applicable, the alleged offender is nevertheless “present” in a territory under the state’s jurisdiction. In such cases, even if the crime was not committed on its territory, and neither the alleged perpetrator nor the victims are its nationals, the state must establish its jurisdiction as long as the alleged perpetrator is present in the territory. This obligation aims to prevent the alleged offender from seeking refuge in a state that has no connection whatsoever with the offence. Or to put it differently, Article 8(3) applies when the crime was committed abroad, and there is no particular connection with the state of the forum. In such a case, the state should exercise universal jurisdiction triggered by the offender’s presence in a territory under its jurisdiction.

France and the United Kingdom (UK) had strong objections to Article 8(3). They argued that existing treaty and customary law did not recognize the obligation to establish jurisdiction with regard to all crimes covered by the Convention on the basis of the perpetrator’s mere presence. Thus, they demanded more flexibility on the scope and application of Article 8(3), especially concerning the presence requirement. On the opposite side of the debate, most delegations had serious concerns over departing from the agreed language and making compromises that could undermine the authority and scope of the obligations to establish and exercise jurisdiction. Accordingly, the Working Group 1 of the Conference had before it the challenge of finding a solution that could bridge these two camps and preserve the consensual adoption of the entire Convention without a vote. The result was a litany of successive proposals tabled for discussion:

(1) replace the verb “shall” with “may” in Article 8(3) (proposed by France);

(2) replace the word “present” with “habitually resident” in Article 8(3) (proposed by France);

(3) leave the language of Article 8(3) untouched but allow states parties to file, with restrictions, a “declaration” stating that they will interpret the word “present” in Article 8(3) as meaning habitual residence (proposed by France and the UK, and substantially reformulated by Austria);

(4) adding the following chapeau to Article 8: “Pursuant to its obligations under international law and for the purposes of this Convention” (proposed by the Chair of the Working Group 1, Professor Vaios Koutroulis);

(5) deletion of the entire Article 8 (proposed by France).

All these proposals were received with mixed responses by the delegations, and none of them was deemed acceptable by both camps. At this point, the negotiations reached an impasse. It was already late in the day, and delegates were exhausted. The chances of getting a consensual solution were slipping away. To move the discussions forward, the Democratic Republic of the Congo (DRC) suggested allowing state parties to make a reservation with limited scope to Article 8(3). The idea was that other states could oppose this reservation should they wish, irrespective of its permissibility under the Convention.

France, the UK, and numerous other states agreed with the DRC’s proposal. Yet, the Argentinean delegation expressed concerns about a reservation-based solution. They argued that allowing a reservation at the heart of the jurisdiction clause of the Convention could go against the object and purpose of the treaty and could send a wrong message in terms of fighting impunity. As an alternative way out, Argentina proposed rephrasing Article 8(3). After incorporating input from the Austrian delegation, the Argentinean proposal of language to Article 8(3) reads as follow: “For the purposes of this Convention, each State Party shall, in accordance with international law, take measures as may be necessary to establish its jurisdiction over the crimes to which this Convention applies.” Although Canada, Finland, Hungary, Malawi, Australia, and South Africa were open to this proposal, the DRC expressed concerns, stressing the importance of not departing from the agreed language and noting that the expression “in accordance with international law” was too broad and vague.

By then, the fatigue in the room was even more evident. As a means to move forward and not block the consensus, Argentina accepted the DRC’s reservation-based solution, which had already been agreed upon by all other delegations. Therefore, while keeping intact the language of Article 8(3) as set out in the zero draft and quoted above, a new third paragraph was introduced into Article 92 (on reservations). After a long discussion on the drafting of this paragraph, all delegations agreed with the following text:

A State may, at the time of signature, ratification, acceptance or approval of or accession to this Convention, formulate a reservation, for renewable periods of three years, based on grounds existing in its domestic law and in accordance with its obligations under international law, limiting the establishment of its jurisdiction under article 8, paragraph 3.

Pursuant to negotiations at Working Group 1 on the drafting of Article 92(3), the use of the word “limiting” in this provision was intended to indicate that a state party cannot, by reservation, entirely exclude the obligation under Article 8(3) to establish universal jurisdiction. Arguably, all that is allowed is for states that wish to do so to limit its scope, including by using the habitual residence of the suspect as the basis of jurisdiction instead of their mere presence.

Data Protection

Protection of personal data also became a divisive issue during the Conference, due to the obligation to carry out international transfers of data in the cooperation channels established by the Convention. The issue triggered a schism between the European and non-European delegations regarding Article 16 of the Convention, originally titled “Protection of information and evidence”.

The members of the European Union (EU), spearheaded especially by Germany and France, stressed that the Convention should reflect the 2016 General Data Protection Regulation. EU members indicated that the absence of a provision sufficiently in alignment with EU law would prevent them from joining the treaty. For this purpose, France and Germany suggested copy-pasting in the Convention the entire text of Article 14 of the 2021 Second Additional Protocol to the Budapest Convention (Second Additional Protocol). They pointed out that, besides the signature by 27 members of the Council of Europe, nine non-European states signed this treaty.

On the other hand, non-European states, including South Korea, Malawi, South Africa, Australia, Canada, and the USA, expressed concerns over including overly prescriptive language on data protection in the Convention. Accordingly, most delegations rejected the Franco-German suggestion of replicating Article 14 of the Second Additional Protocol, indicating that this treaty and the present Convention have different scopes and cover distinct aspects of data protection. The lengthy, detailed, and prescriptive text of Article 14, with approximately 2000 words, was another issue.

Malawi argued for a compromise that was ultimately accepted by all delegations: instead of the inclusion of numerous detailed obligations and safeguards in the text, the Convention should have broader language reflecting the main data protection principles. This solution could accommodate the detailed and stringent EU law, without imposing the latter upon the world via the Convention. EU members accepted this proposal but insisted on the inclusion of the following Article 16(10) in the Convention: “The requested State Party shall not be obliged to transfer personal data if the domestic law applicable to it prohibits transfer or if it has reason to assume that the legitimate interests of the person concerned would be adversely affected by such transfer.”

In the end, Article 16 of the Convention, newly titled “Use and protection of personal data”, was adopted with this tenth paragraph. It remains to be seen if Article 16 will prevent states from joining the Convention, either because its content is too rigorous or because it is not rigorous enough. For the states that ratify the Convention, one wonders about the effect of Article 16(10)’s wide margin to refuse requests for transfer of personal data in the Convention’s implementation.

Rights of Victims

The Convention constitutes a milestone towards strengthening the international protection of victims. Its Article 81 contains a definition of victims mirroring Rule 85 of the Rules of Procedure and Evidence of the International Criminal Court. Article 82 obliges states parties to “take appropriate measures within its means to provide effective protection from potential retaliation or intimidation” of victims, witnesses, experts and other persons. These protective measures include relocation and non-disclosure of information that could reveal the person’s identity. Finally, Article 83 ensures to victims the right to reparation and participation in criminal proceedings.

However, these three provisions triggered intense debate and negotiations between state delegates and between them and NGOs. Some delegations, including France, UK, Austria, and Germany, were adamant on the need to ensure significant room for flexibility in the implementation of the provisions on victims as well as a clear indication that Articles 81-83 were limited to the restricted scope of the Convention and could not serve as a legal basis to compel states parties to carry out comprehensive changes in their domestic legal systems. In other words, the Convention should adapt to existing national laws and not the other way around.

The result was a complex series of compromises, evinced especially by the repeated insertion of expressions such as “for the purposes of this Convention”, “in accordance with domestic law”, “as appropriate”, or analogous language to the same effect, in the text of Articles 81-83. The inclusion of these multiple caveats in different parts of the provisions led to an intricate language that may give rise to complications and uncertainty during their interpretation and application. This was the price for the consensus adoption of Articles 81-83 in the Convention.

Concluding Remarks  

Regarding whether the gaps in treaty law have been closed, an aspect on which this Convention was most eagerly awaited is the aut dedere aut judicare obligation (the obligation to extradite or prosecute). Through this obligation, the duty to cooperate in combating impunity for international crimes is given effect. It is to be welcomed that the Convention includes Article 14 on aut dedere aut judicare, modelled on the Hague Formula as enshrined in Article 7 of the CAT and as interpreted by the International Court of Justice in the Hissène Habré case (see here, paras. 94-95). In its principle, the instituted aut dedere aut judicare obligation is clear in that the obligation to prosecute is triggered irrespective of the existence of a prior request for the extradition of the suspect. Likewise, the priority of prosecution vis-à-vis extradition is established, in the sense that, as the ICJ stated (see here, para. 95), “the choice between extradition or submission for prosecution […] does not mean that the two alternatives are to be given the same weight. Extradition is an option offered to the State […], whereas prosecution is an international obligation.”

In its system, the Convention also provides for procedural requirements to facilitate the obligation to prosecute, such as criminalization in domestic law (Art. 7), the establishment of domestic jurisdiction (Art. 8), and the preliminary measures (Art. 13). As the ICJ concluded regarding a similar set of provisions in the CAT, “[t]hese obligations, taken as a whole, may be regarded as elements of a single conventional mechanism aimed at preventing suspects from escaping the consequences of their criminal responsibility, if proven” (see here, para. 91). In addition, procedural tools to facilitate extradition are established if the state wishes to discharge its obligation to prosecute by extraditing the person concerned (see Arts. 49-65). Finally, the detailed provisions on mutual legal assistance provide a solid and indispensable model for cooperation between states (Arts. 23-48).

Now it is up to the states to join this system. If the Convention gains only a limited number of state parties, the legal gap that originally motivated the establishment of the MLA Initiative would remain open. After all, having a treaty without worldwide adherence would be moot for achieving the purpose of fostering cooperation on criminal matters at a global level.

Photo: ‘Baroque ceiling frescoes of Cathedral in Ljubljana, Slovenia. Work of Italian master Giulio Quaglio in 1703–1706 and later 1721 1723’ (Petar Milošević, 2021)