Europe and the World: A law review The integration paradox: an ILC view on the EU contribution to the codification and development of rules of general international law

The contribution of international organisations (IOs) to the process of identification, codification or development of rules of general international law is one still enveloped in a measure of mystery. While the United Nations (UN) International Law Commission (ILC) has long relied on IOs’ practice, and the practice of States within IOs, in carrying out its work, it has only recently begun to address the role that IOs might play in the formation or expression of rules of customary international law, jus cogens or general principles of law. In this process, the European Union as evidence of existing or emerging rules of general application. This article looks at the effects of this ‘integration paradox’ in the ILC’s work by reviewing the references to, and the use of, EU practice in eight distinct codification projects, combined with interviews with ILC members and EU officials. It provides an ILC (outside) view on the relevance of EU practice for the identification, codification, and development of rules of general international law. This view, in turn, has implications for the operation of an EU (inside) foreign policy objective: its ambition to contribute to the development of international law, as expressed in Article 3(5) of the Treaty on European Union (TEU).

emerging 'global rule maker' 8 with an ambition to 'transform rather than to simply preserve the existing [international] system', 9 to 'shape the global future' 10 and to defend and extend international norms. 11 EU legal scholarship has likewise shed light on the different ways through which the EU may 'shape' international law, ranging from its active participation in international agreements or IOs, to its very existence. 12 This article addresses this EU foreign policy objective from the viewpoint of the 'distinctive community' 13 to which it relates, in other words, from an outside perspective. By reviewing how the ILC has accounted for EU practice in eight distinct codification projects, 14 the article demonstrates that, as far as rules of general international law are concerned, EU integration is a double-edged sword. On the one hand, the transfer of powers to the organisation has rendered it particularly well suited to 'shape' international rules. As its competences expand, the EU will often have relevant practice to showcase in different fields. On the other hand, the unique features of EU integration, coupled with the ILC's ambivalence regarding IOs' role in the formation or expression of rules of general international law, explain this body's reservations about drafting general rules modelled on EU-related practice. While the EU's 'exceptionalism' has not prevented the ILC from occasionally proposing rules grounded on, or inspired by, the EU legal system or its international practice -where little practice or precedent were available to support the ILC in its work, or where EU-related practices were aligned with the normative aims of a particular project -these attempts at generalisation have often been met with strong objections by States at the Sixth Committee. As a whole, EU practice has, more often than not, been deemed too 'exceptional' to serve as evidence for the codification of rules of general application. Additional factors such as ILC members' familiarity with the EU legal system and the availability (and intelligibility) of EU practice also contribute to this body's approach to the EU.
The article is structured as follows: Section 2 draws from an EU law viewpoint (the inside) to frame EU statements on the ILC work as a manifestation of its ambition to contribute to the development of international law. Sections 3 and 4 contextualise this ambition against the ILC's (outside) understanding of the relevance of IOs' practice in general, and that of the EU specifically, for the identification or development of rules of general international law. Section 3 outlines the working methods of the ILC and discusses how the views and practice of IOs make their way into ILC codification projects. Special attention is accorded to the 2018 Conclusions on Customary International Law (CIL), where a debate emerged concerning the ability of IOs to contribute to the formation or expression of rules of custom. Section 4 applies these considerations to the use of EU practice by the ILC, surveying the references to the EU in eight distinct codification projects. The sample focuses on projects regarding which the EU has made statements at the Sixth Committee (where the ILC annual report is discussed) and which have been adopted by the ILC on second reading. 15 The analysis is further informed by interviews with ILC 8  that international law is often described as 'a set of rules that arises from the practices and usages of a distinctive community').
14 These include (i) the draft articles on most-favoured-nation clauses, ILC Annual Report (1978) A/33/10, chapter II, para 74 ('MFN articles'); (ii) the draft articles on the law of treaties between states and international organisations or between international organisations, ILC Annual Report (1982) A/37/10, chapter II, para 63 ('1982 draft VCLT-IO'); (iii) the draft articles on the responsibility of international organisations, ILC Annual Report (2011) A/66/10, chapter V, para 87 ('ARIO'); (iv) the draft articles on the expulsion of aliens, ILC Annual Report (2014) A/69/10, chapter IV, para 45 ('EoA articles'); (v) the draft articles on the protection of persons in the event of disasters, ILC Annual Report (2016) A/71/10, chapter IV, para 48 ('PPED articles'); (vi) the conclusions on subsequent agreements and subsequent practice in relation to the interpretation of treaties, ILC Annual Report (2018) A/73/10, chapter IV, para 51 ('SASP Conclusions'); (vii) the CIL Conclusions (2018); and (viii) the draft articles on the prevention and punishment of crimes against humanity, ILC Annual Report (2019) A/74/10, chapter IV, para 45 ('CAH articles'). 15 A few additional points on methodology are in order. First, although the EU has submitted statements on other ILC projects and occasional references to these are occasionally made throughout the text, the analysis excludes topics still on first reading at the time of writing (e.g. the Guide on the provisional application of treaties, ILC Annual Report (2018) A/73/10, chap. VII, para 79 ('Guide on PA'), as well as those where no reference to the EU is made in the final set of articles, notwithstanding its statements members and EU officials, carried out between 2019 and 2020. Section 5 offers some conclusions on how the relevance of EU practice to the codification and development of rules of general international law is seen from the outside, and why integration has rendered it both well suited and, paradoxically, too exceptional to fulfil the ambition underlying Article 3(5) TEU as far as rules of general international law are concerned.
2. An inside objective: the EU commitment to contribute to the development of international law and the role of the ILC therein Article 3(5) TEU is a legally binding 'promotional norm'. 16 As argued by Larik, these types of norms impose on the EU institutions and, by virtue of Member States' obligation of loyalty under the Treaties, on Member States themselves, 17 an obligation to apply their best efforts in the pursuit of (in casu) the development of international law, as part of the promotion of a 'European common good'. 18 What the pursuit of this objective entails, however, is far from clear.
CJEU case law has so far offered little guidance regarding the content of this foreign policy objective. As noted by Milano, the Court 'is generally concerned with upholding and guaranteeing the consistency and coherence of the EU legal order, rather than contributing to the development of international law or projecting the image of the EU as a Völkerrechtsfreundlich actor'. 19 The Court's reasoning, however, often betrays an understanding of EU law as either part of a larger trend towards a progressive development of international rules 20 or as a more developed form of these rules. 21 In turn, the efforts to be deployed in the fulfilment of this ambition appear to be boundless, at first sight limited only by the organisation's competences and by the external conditions in which it exercises these powers.
The preparation and delivery of EU statements concerning ILC codification projects are part and parcel of this process. In full exercise of its UNGA observer status, 22 the EU has since 1975 participated -somewhat erratically 23 -in the debates concerning the ILC work. It has submitted statements on the on the project (e.g. the articles on the effect of armed conflicts on treaties, ILC Annual Report (2011) A/66/10, chapter VI, para 89). Second, the absence of EU statements on a particular ILC topic does not necessarily mean that its practice is not accounted for by the ILC. Third, focusing the analysis on the final output of an ILC project does not imply that EU practice was not analysed in the context of Special Rapporteur (SR) reports or ILC debates, having nevertheless been omitted from the final set of articles, conclusions or guidelines. To reduce the object of analysis, however, a choice was made to focus on topics where the EU expressed interest in contributing to the ILC work and on the final texts as adopted by the ILC. Finally, 'contribution' is measured here in a formal and narrow sense, by the number of references to EU practice in the ILC work. 16 Larik (n 7) 168. See also Bart Van Vooren, Steven Blockmans and Jan Wouters (eds), The EU's Role in Global Governance: The Legal Dimension (OUP 2013) 1 (noting that '[t]his role for the Union whereby it "stabilizes" the world and "points the way ahead" is not merely a moral imperative proclaimed by political leaders, but has found its way into EU primary law as a legally binding obligation'). 17  codification and development of rules of general international law in areas as diverse as the responsibility of IOs or the prevention and punishment of crimes against humanity (CAH). Its statements are mostly prepared by the Legal Service of the European Commission and discussed within the Council of the EU Working Party on Public International law (COJUR), before being sent to the EU Delegation to the UN in New York. 24 The choice of whether or not to make a statement follows not only an assessment of the EU competence and relevant practice on the topic, but also of the EU's own perception of the ILC's authority and of the impact of its work on EU external relations. It involves a 'fingertip feel' on whether the matter should be left to the sole consideration of (Member) States -who retain their right to make individual statements -or also include an EU dimension. 25 While these EU statements can hardly be qualified as an organisational practice directly contributing to 'the development of international law', they are an important expression of this practice. The EU's participation in these debates is a relevant part of this process in at least three ways. First, by coordinating (internally) a position on questions of public international law among its Member States at the COJUR level, the EU operates as a platform which can reinforce the importance of specific questions of international law or allow for Member States' views to converge around a common opinio juris. 26 Second, by participating (externally) in Sixth Committee debates on the ILC work, the EU presence (or visibility) is an expression of institutional practice of how an organisation operates in a multilateral context, alongside its Member States. 27 Third, EU statements as such can assist in the identification of rules of international law. They convey evidence of the organisation's practice and are an expression of its views on the existence or emergence of international rules. 28 As far as gauging how 'significant' the EU contribution is to the ILC work, this is far from consensual. There seems to be some agreement that the ILC 'may form a venue for EU influence on international law-making', 29 a fact which in part explains why, within the Sixth Committee, 'the EU seems to attach more importance to providing input to the work of the ILC'. 30 In practice, however, the impression is that the EU 'influence' on the outcome of the ILC work, while present, is relatively marginal. 31 The following sections will address this question from the viewpoint of the ILC. They first outline the working methods and mandate of the ILC, and how this body has relied on (and conceptualised the legal relevance of) the practice of IOs in its work, before distinguishing the EU from this larger category of institutional actors.

The ILC working methods and the role of international organisations
In broad strokes, the ILC mandate can be defined as that of carrying out research and preparing draft conventions which either systematise and clarify well-established rules of general international law, or which propose rules on matters regarding which international practice is considered to be insufficiently developed. 32 Both dimensions of this (often criticised) binary distinction between codification and (progressive) development of international law 33 follow a relatively similar methodology: that of surveying existing practice -namely 'the texts of laws, decrees, judicial decisions, treaties, diplomatic 24  correspondence and other documents relevant to the topic' 34 -and distilling from it rules which reflect either a well-ingrained practice or an evolving and emerging one. 35 ILC members therefore see their mandate as one of 'bringing a degree of consistency to international law' by 'codifying custom' or proposing rules which advance the values of this legal system and develop its vocabulary. 36 The output of this work, in turn, operates as a subsidiary source of the very same system it aims to clarify and advance. 37 There is an inherently interpretative and selective dimension to this effort. 38 It is present in the act of surveying 'the whole field of international law' and selecting the topics therein the codification of which is 'necessary and desirable'; 39 in the identification of practice that is not only available but also interpreted as relevant (or legally significant) both by the system and by those making the selection; and in the act of organising it in a manner which proves or disproves the 'extent of agreement on each point in the practice of States and in doctrine'. 40 Perhaps in recognition of the subjective dimension of this exercise, the ILC statute requires that this body be comprised of qualified persons representing 'the main forms of civilization and of the principal legal systems of the world' 41 and that it operate in a collegial and analytically rigorous manner, receiving input not only from within, but also from the larger community of international actors, without.
The ILC statute provides for a number of ways for external input to inform the process of gathering, filtering, interpreting and codifying international norms. The ILC may consult with UN organs, official or non-official international or national organisations, and scientific institutions and individual experts; 42 it may submit requests for information from governments, organs, specialised agencies and official bodies; 43 and it must take into account the comments received. 44 Each topic is subject to a relatively long-term discussion among ILC members, as well as with States and other international actors represented at the Sixth Committee. In addition, information is also gathered through formalised exchanges between the ILC and a relatively fixed set of juridical bodies of mostly regional intergovernmental organisations. These include the Inter-American Juridical Committee (IAJC) and the Council of Europe Committee of Legal Advisers on Public International law (CAHDI), 45 which present their own work and practice of relevance to ILC projects during ILC annual meetings in Geneva.
Beyond a consultative function, the role that IOs and their practice play in the process of formation or expression of general rules has not been absent from the ILC work. While it is a truism to say that international law is State centric as a legal system, 46 the ILC has often relied on IOs' practice when discussing rules of general international law. 47 This was necessarily the case where projects specifically regulated the legal position of IOs, such as the 1982 draft articles on the law of treaties between States and IOs or between IOs (VCLT-IO) and the 2011 draft articles on the responsibility of IOs (ARIO). Where ILC projects adopted instead a 'combined' codification method, developing rules on a particular topic potentially relevant to both States and IOs, 48 the reliance on the practice of States in the context of IOs, or of IOs 'as such' in the exercise of their competences in the field subject to regulation, has 34 Art 19(2) ILC statute. 35 See, inter alia, commentary to conclusion 7, para 4 and commentary to conclusion 18, para 5, Jus Cogens Conclusions -first reading (referring to IOs' obligation to 'act, within their respective mandates and when permitted to do so under international law' to bring to an end serious breaches of jus cogens norms); arts 4(2) and 14, CAH articles (referring to States' obligations to act through IOs in the prevention and punishment of CAH). also been of relevance. For instance, in drafting rules on the protection of persons in the event of disasters, Valencia-Ospina, Special Rapporteur on the project, relied significantly on the practice of the International Federation of the Red Cross and Red Crescent Societies (IFRC) in the provision of international disaster relief and assistance. 49 In drafting rules on the expulsion of aliens, in turn, Maurice Kamto turned to EU directives as a source of 'inspiration' to formulate general rules governing the conduct of States in this field. 50 The exact weight and legal relevance accorded to the practice of IOs in this process, however, is difficult to gauge. This question assumed some prominence in the debates surrounding the CIL Conclusions and has been dealt with, more indirectly, in a number of topics recently addressed by the ILC, including on 'peremptory norms of general international law (jus cogens)' and the preliminary debates on 'general principles of law'. 51 Combined, these projects reflect a particular understanding, within the ILC, and arguably international law at large, about the role that IOs and their practice might play in the co-creation of rules or principles of general international law and, therefore, their ability to contribute to this system and, as a consequence, to the work of the ILC.

The practice of international organisations in the formation or expression of rules of general international law
In 2018 the ILC adopted, on second reading, a set of 16 conclusions on the identification of CIL, which included conclusion 4(2), stipulating that '[i]n certain cases, the practice of international organizations also contributes to the formation, or expression, of rules of customary international law'. Simple as it may seem, and perhaps because of it, this conclusion has engendered much debate. This debate pitted two visions on the role of IOs within the fabric of international law against each other: one which denies that IOs might have a role in the formation of rules of international law, and thus views them mostly as 'objects to be regulated by general international rules developed by states'; 52 and one that asserts that IOs 'can and do play' a role in this process. 53 The opinions on conclusion 4(2) shared at the Sixth Committee varied from seeing it as 'too limited' to 'sufficiently reflect the growing participation of universal as well as regional [IOs] in the international relations' 54 to seeing it as a progressive development of international law focused on the 'limited experience' of the EU. 55 Within the ILC, a similarly wide range of views was expressed: Some members …were of the view that the practice of international organizations was not to be taken into account in the process of identification of rules of customary international law. Other members considered that the practice of international organizations was only pertinent to the extent it reflected the practice of States. Some other members, however, agreed with the Special Rapporteur that the practice of international organizations as such could be relevant to the establishment of customary rules, particularly in regards to certain fields of activity within the mandates of those organizations. 56 What the resulting conclusions tell us about the ILC views, as informed by the views of States and the limited number of IOs which pronounced themselves on this project, is that the practice of IOs may be legally significant to the formation or expression of custom either (i) indirectly (when their conduct 49 See commentary to art 3 (use of terms), paras 3 and 13; commentary to art 4 (human dignity), para 3; commentary to art 6 (humanitarian principles), paras 6-7; commentary to art 17 (duty of the affected States to seek external assistance), para 5; commentary to art 15 (facilitation of external assistance), para 3, PPED articles. reflects the collective 'practice and convictions' of States or 'catalyses' or 'prompts' State practice) 57 or (ii) in 'certain cases' directly (as autonomous and independent 'actors in their own right'). 58 Both cases are seen through the lens of the State as the primary unit. In the first model, IOs are fully transparent and only indirectly relevant as structures for the expression of the views of States. 59 In the second, they are seemingly opaque but only (in the 'certain cases') where their conduct corresponds to the exercise of State-transferred or State-like powers. 60 The universe of IOs that can contribute directly to the formation or expression of rules of custom, therefore, is significantly smaller than the universe of IOs in existence. 61 The weight accorded to their practice, in turn, seems to depend as well on additional criteria, including the size of an IO's membership, 'the nature of the organization' and 'whether the conduct is consonant with that of the member States of the organization'. 62 This last requirement, in particular, reflects the ILC's reservations about IOs' ability to operate as autonomous 'jus generative forces', 63 in lieu of or alongside their member States.
The 'ambivalence' that Blokker has ascribed to the ILC's approach to IOs in its CIL Conclusions is transversal to the ILC projects that have followed, notably those dealing with jus cogens norms or general principles of international law. 64 If the ILC concluded that IOs might contribute to the formation or expression of rules of custom, it indicated the contrary as far as peremptory norms of general international law are concerned. 65 Conclusion 7 of the ILC work on jus cogens, as adopted on first reading (2019), notes that '[w]hile the positions of [actors other than States] may be relevant in providing context and for assessing acceptance and recognition [of jus cogens norms] …these positions cannot, in and of themselves, form part of such acceptance and recognition'. 66 The commentary to the text clarifies that, notwithstanding minority views to the contrary, 'the current state of international law retains States as the entities whose acceptance and recognition is relevant'. 67 In the ILC's work on general principles of law, ongoing at the time of writing, Special Rapporteur Vázquez-Bermúdez has also drawn the ILC's attention to the need to discuss 'whether international organizations could also contribute to the formation of general principles of law'. 68 Ultimately, what these ILC projects show is a growing, if 'modest', recognition of the relevance of IOs in the fabric of international law. 69 ILC members have noted that they do not want to be viewed as downplaying the role of IOs, but still approach this role with the reserved caution warranted by the State-centric system they have been appointed to serve. 70 While discussions have shifted from whether IOs might contribute to the formation or development of rules of (customary) international law to how or when they do so, 71 their participation in this system remains secondary to that of States.
The question then becomes: where does the EU fit in all this? Conclusion 4(2) of the CIL project seems at first sight rather felicitous for the organisation. 72 The EU is the single actor referred to in the commentary to this conclusion as a 'clear-cut' example where 'the relevance of practice is difficult to deny'. 73 If relevance is gauged by the transfer of powers to the organisation, EU integration places it at a particular advantage. Its growing competences mean that the EU will often have relevant practice to showcase in different fields of activity, from the conclusion of international agreements to the protection of persons from natural or manmade disasters. At the same time, this degree of integration also renders it rather 'unique'. How does this 'uniqueness' affect the reliance on EU practice in the identification, codification or development of rules of general international law? The following section addresses this question by reviewing the use of EU practice in eight distinct codification projects. It shows how the singularities of the EU system are both well fitted to contribute to international law, and often too exceptional to do so.

The EU and its statements on the ILC work
For a body reliant on information from States and IOs for the adequate pursuit of its mandate, active participation of the EU in Sixth Committee debates is far from unwelcome; quite to the contrary, the ILC has often regretted the limited engagement of States and IOs with its work. 74 As far as EU participation is concerned, the dominant perception seems to be that, when the EU speaks at the Sixth Committee, it voices a well-crafted and intensely coordinated position, reflecting the compromise agreement between its Member States. 75 But while there is an overall convergence between the views expressed by the EU and those of its members, 76 with the latter at times referring first to the EU statements before adding to them, 77 this convergence is not always absolute. 78 When the EU speaks, its position is mostly accounted for as (just) 'one more view', not that of (now) 27 States, and as such as the view of an IO. 79 In a process where the views accounted for are, first and foremost, those of States, 80 some ILC members have noted that the shift from statements voiced by a State on behalf of the EU to those of the EU qua organisation 81 brought with it a measure of weakness of no longer being dressed in the cloak of statehood. 82 The substance of EU statements, in turn, has received mixed reviews. These have varied from detailed comments on specific draft articles to rather vague remarks on the objectives pursued by an ILC project. 83 From the ILC side, EU statements have been seen as everything from 'extremely useful' 84 or 'very constructive…not just formalistic but also substantive', 85 to generally 'useful' 86 or 'of an extraordinary emptiness' reflecting a 'pathetic input' at times left unnoticed. 87 While some ILC members 72 CIL Conclusions, commentary to conclusion 4(2), para 6. 73 A/CN.4/682, Third report on identification of customary international law (2015), para 77; A/CN.4/695 and Add.1, Fourth Report on the identification of customary international law (2016), para 20. 74  The Integration Paradox 10 have noted that they did not experience particular difficulties in understanding the intricacies of EU practice of relevance to a particular project, others have noted that the complexity of the EU legal system is far from helpful. 88 Some ILC members noted that a greater investment in making EU practice of relevance to the ILC work more readily 'available' would assist in this process. The former habit of publishing a survey of EU practice of relevance to international law in specialised issues of academic journals was remarked as a particularly helpful one, and one which might be revisited. 89 The conceptualisation of the EU and its legal system have likewise varied significantly in ILC projects and among the ILC membership. The EU has been described as 'a hybrid union of States' similar to an intergovernmental organisation, 90 an 'obviously unique phenomenon', 91 an IO with 'some supranationalisms', 92 an 'interesting innovation' or a legal 'experiment' in governance, 93 or something akin to 'a federation of states'. 94 Under the draft rules on the expulsion of aliens, EU practice is understood as the practice of 'a community of states'. 95 In the draft articles on the protection of persons in the event of disasters, it is interchangeably used as the expression of the practice of a regional (integration) organisation or as the regional practice of its members. 96 The EU legal order, in turn, has been described as everything from a 'regional order' 97 or a 'special regime', 98 to 'a peculiar legal system' 99 akin to 'the domestic law of a large federal state', 100 or a 'sub-system' of international law. 101 While in more recent ILC projects this 'difference' is approached in a rather routine fashion, being acknowledged somewhat matter-of-factly, 102 it gave rise to some debate in ILC projects in the 1970s. 103 This conceptualisation, and ambivalence, have not been without consequence. 104 As the next section shows, the EU's 'exceptionalism' has warranted distinct results as far as the codification of rules of general international law is concerned.

References to and use of EU practice in ILC projects
and EU primary law 107 as well as secondary law. 108 In addition, references include acts of EU institutions such as Council Conclusions 109 or European Commission policy and legislative proposals, 110 unilateral acts or statements, 111 EU practice as a participant in IOs, including its statements therein, 112 EU external action in monitoring missions, 113 CJEU case law 114 and, significantly, references to cases in which the EU has acted as a party, namely within World Trade Organization (WTO) dispute settlement proceedings. 115 The majority of these references are included in footnotes. Express references to the EU in the commentaries to texts adopted by the ILC on second reading vary between a single reference in the articles on the protection of persons in the event of disasters (2016), 116 the CIL Conclusions (2018) 117 or the draft CAH articles (2019), 118 to a total of 12 references in the commentaries to the ARIO (2011). 119 There appears to be little correlation between the number of EU statements on a particular topic and the number of references to its practice ultimately included in ILC drafts. For instance, while the ILC articles on the expulsion of aliens (2014) relied at first heavily on EU practice 120 and the EU made statements at the Sixth Committee on five distinct occasions, 121 the final draft refers to EU practice in three footnotes and the commentary to two draft articles, 122 a fact not left unnoticed by the EU delegation. 123 What becomes clear from the review of the use of EU practice in the ILC work is the degree to which this practice is perceived as exceptional, and the different results that this 'otherness' or 'exceptionalism' have produced. EU practice has been used as evidence for the codification of general rules in situations where (alternative) practice and precedent were lacking, as a subsidiary source confirming the established practice of States or IOs, as the illustration of a legal regime comprising more detailed rules, or as the example of a 'unique' case inapt to serve as evidence of or inspiration for the formulation of rules of general application.
The first of these approaches has mostly been reserved for the codification of rules regulating the legal position of IOs. For instance, the inclusion in the 1982 draft VCLT-IO of a rule on reservations to treaties by IOs notes that the majority of 'precedents concern the European Economic Community' (EEC). 124 In the ARIO, a text largely cast as an exercise in progressive development, 125 the inclusion of a rule recognising IOs' right to invoke responsibility for the breach of obligations owed to the international community as a whole was grounded on the caveat that 'practice in this regard is not very indicative …The most significant practice appears to be that of the European Union.' 126 In these cases, EU practice, exceptional or not, was deemed both useful and sufficient to serve as evidence for general rules. Beyond IO-specific rules, the clearest example of the use of EU practice as evidence of the existence of a general rule is perhaps that found in the 1978 draft articles on most-favoured-nation (MFN) clauses in trade agreements, a matter squarely within EU exclusive competence. Here, the EEC-ACP Lomé Convention (together with the 1975 Bangkok Preferential Trade Agreement among Asian developing countries) was used as evidence of a rule excluding preferences granted between developing countries inter se from the operation of MFN clauses in multilateral treaties concluded between developed and developing countries. 127 Unsurprisingly, EU practice has also featured as subsidiary evidence in the identification of general rules. For instance, CJEU case law has been used to confirm the view that Article 31 VCLT has customary law status, 128 and Council conclusions and European Commission acts have been used as evidence of the 'widespread practice of States reflecting their commitment to reduce the risk of disasters'. 129 On other occasions, the singular features of the EU legal system and international practice have instead been referred to as illustrations of a more detailed system of rules. In the CAH articles, for instance, the EU Charter of Fundamental Rights serves as an example of a legal instrument with 'more specific standards binding upon States', as do the criteria for dealing with multiple requests for surrender found in the European Arrest Warrant Framework Decision. 130 The integration of these rules into general norms, however, has been addressed with caution. Rules which refer 'in too exclusive a manner to a case as special as' the EU have often been rejected for lack of support in general practice. 131 Both States at the Sixth Committee and ILC members themselves have at different points in time expressed reservations about 'attempt[s] to formulate provisions valid for only one organization, whose character was basically different from that of the great majority of international organizations'. 132 Most emblematically, the ILC rejected the EU's request for the codification of special rules of attribution for regional economic integration organisations (REIOs) -a 'code-name' for the EU 133 -within the ARIO on the basis 'of the practical relevance [of such rules] only for a limited number of international organizations'. 134 Concerns about the reliance on EU practice as evidence of general rules were also expressed in the debates on the draft articles on the expulsion of aliens. The United States delegation, in particular, cautioned the ILC not to 'seek to codify new rights, or to import concepts from such regional bodies as the European Commission or the European Court of Human Rights'. 135 The choice not to draft rules of general application along the lines of EU practice, however, has also at times reflected larger normative concerns about the implications of EU practice for specific principles of international law, or a judgment on the desirability or lack thereof of EU integration. The latter was most prominent in the early years of the EEC and the ILC debates on the draft articles on MFN clauses. In this context, the ILC dismissed an EEC-backed rule recognising an automatic exemption of intra-custom unions' benefits from the reach of MFN clauses on account of the absence of 'compelling evidence' of the benefits of regional economic integration among developed nations. 136 Later projects have expressed some concerns about the compatibility of EU treaty practices with the principle of State consent. The inclusion of a reference to the prohibition on invoking the internal rules of the organisation as a justification for a failure to perform a treaty in the ILC guidelines on provisional application (PA) of treaties, for instance, has been cast as a reaction to the EU's practice of defining the scope of PA of mixed agreements rather broadly, by reference to 'matters falling within the Union's competence', to the detriment of the position of EU treaty parties. 137 Recalling that 'general principles are not stated as non-derogable rules', 138 the ILC has often accounted for the EU (and the institutional diversity of IOs more generally) through a variety of codification techniques. These have included 'without prejudice' clauses, 139 clarifications to a project's use of terms, 140 the indication of the non-exhaustive nature of certain provisions, 141 the indication that a project's particular provision 'does not attempt to express a clear-cut view on the issue', 142 deferring the regulation of certain essential questions to the rules of the organisation, 143 and ultimately, the codification of a lex specialis clause. The well-known example of the latter is Article 64 ARIO, the commentary to which refers expressly to the EU. This choice has not been without criticism. Some authors have ventured that the ILC might adopt a more methodological appraisal of the legal relevance of IOs' practice or overcome its reservations to recognising the existence of different rules for different types of organisations. 144 Ultimately, as far as the EU is concerned, the ILC has accounted for its views and its practice with ponderation but reservation, a reservation otherwise extended to IOs at large.

Conclusions
In 'The Logic of Paradox', Graham Priest suggests, rather unorthodoxically, that one should learn to live with certain paradoxes. 145 By explaining how the ILC relies on the practice of IOs in its work, and how it has conceptualised the legal relevance of this practice to the formation and expression of rules of general international law, this article laid bare a paradox of integration, one with which the EU must live when making statements on the work of the ILC.
As the international debate currently stands, the practice of IOs is relevant for the identification, codification or development of rules of general international law in rather distinct ways. It will necessarily be relied upon by the ILC when codifying or developing rules regulating the position of IOs as treaty parties or as respondents for international wrongs. It will also be relevant where IOs are particularly active in a field the ILC seeks to systematise through general rules, be this crisis management or atmospheric protection. Conceptually, the ILC concluded in 2018 that the practice of IOs might co-contribute, 'in