How the Court Tries to Deliver Justice in Common Foreign and Security Policy, Where the Need for Judicial Protection Clashes with the Principles of Conferral and Institutional Balance. Joined Cases C-29/22 P and C-44/22 P KS and KD

Abstract: KS and KD is a judgment where the European Court of Justice finds it has jurisdiction to hear actions for damages caused by the European Union’s (EU) civilian missions or military operations. It is the first time that this has happened. Although not entirely surprising, the finding of jurisdiction contributes to a growing line of case law that has proved controversial as it is in tension with a literal reading of the Treaties, which severely limit the Court’s jurisdiction in this domain. This Insight shows how the case law of the Court on Common Foreign and Security Policy (CFSP) seeks to strike a balance between the principles of conferral and of institutional balance, on the one hand, and of effective judicial protection, on the other hand. The need to find such a balance is made difficult by the peculiar structure of the powers of the Court under the Treaties, which contrasts with the reality of an increased need for judicial protection in the field of foreign policy. The outcome, in KS and KD is a judgment that is in tension with the wording of the Treaties, but which is nonetheless a welcome result as it delivers some justice, and it is in line with the case law on EU CFSP. The Insight also discusses what remains outside the Court’s jurisdiction, namely foreign policy decisions “directly relating to political or strategic choices”.

Keywords: Common Foreign and Security Policy – Common Security and Defence Policy – jurisdiction – human rights responsibility – action for damages – EULEX Kosovo.

I. Introduction

On September 2024, the Grand Chamber of the European Court of Justice delivered its judgment in joined cases KS and KD,[1] affirming jurisdiction, for the first time, over action for damages allegedly caused by the European Union’s (EU) civilian missions or military operations. It also found that, in the context of the EU’s Common Foreign and Security Policy (CFSP), the Court of Justice of the EU (CJEU) courts do not have jurisdiction over decisions directly relating to political or strategic choices.

Neither of those findings is obvious by reading EU Treaties or precedents.[2] Even though it only concerns a procedural point, KS and KD is therefore an important pronouncement. Precedent was not decisive, as the extent of the Court’s jurisdiction in the context of CFSP is unclear (it is now somewhat clearer, although not by much).[3] The fact that the Court has jurisdiction has significant real-world consequences: the EU has operated over 40 missions,[4] and many people could now seek compensation if they allege that the EU violated their rights. The outcome of the case is also interesting from the perspective of legal reasoning. The result reached by the Court is in tension with the wording of the Treaties and the judgment is imbued with the language of effectiveness and systematic interpretation.[5] Arguably, this non-literal reading is necessary to deliver justice: like many others CFSP cases, KS and KD underscores a tension, highlighted by the increase in EU foreign policy action, between principles of EU law.[6] On the one hand, the principles of conferral and of institutional balance  restrict the Court’s jurisdiction in foreign policy. On the other hand, the principle of effective judicial protection calls for access to justice. With its classic “constitutional” approach,[7] the Court does its best to deliver justice by balancing those principles, perhaps stretching, but not shaking, the yoke that the principle of conferral places on it. This principle also limits the paramountcy of fundamental rights, disappointingly for some commentators,[8] and contrary to what, in essence, Advocate General Ćapeta had suggested in her Opinion in this case. Yet, the Insight argues that the outcome of the judgment is defensible because, among the various options that the silences and open-ended nature of the wording of the Treaties allowed (including the one proposed by the Advocate General), the Court chooses one that closely adheres to its (reinterpretation of its own) precedents, and because it strikes a fair balance between the principles of institutional balance and of conferral on the one hand, and to effective judicial protection on the other hand. When the Court introduces the vague formula that it will not review decisions “directly relating to a political or strategic choice of the Union”, the tension with the wording of the Treaties, which contain no such a sentence, is strong, but the Court does not go too far. First, arguably, a similar conclusion should have been reached even in the absence of arts 24 TEU and 275 TFEU, because the principle of institutional balance and separation of powers may be interpreted as requiring it.[9] Second, it shall be recalled that KS and KD is about procedure. The Court says nothing about the merit of the applicants’ claims. The ECJ limits itself, as it did in Kadi,[10] Venezuela,[11] and Bank Refah,[12] to offer some “justice” in the procedure as a matter of principle, leaving the outcome on the substance to a future, factual determination, by the General Court.

The judgment is relevant not only for the EU constitutional order, but also for the international protection of fundamental rights more broadly. The judgment may be read as implicitly confirming what the ECJ had found in Opinion 2/13:[13] the structure of the Court’s powers under the EU Treaties, which restrict the Court’s judicial review to only some foreign policy acts,[14] is one of the “essential characteristics” of the EU legal order that must be protected if the EU is to accede to the European Convention for Human Rights.[15]

In Section II, this Insight outlines the factual and legal context of the judgment. In Section III, it summarises the decision and its rationale, and highlights the points on which it differs from the Opinion of the AG. Section IV discusses two issues: firstly, the Court’s reasoning shows its approach to precedent; secondly, future judgments will need to clarify what decisions are directly related to the EU’s strategic or political choice – and are therefore not reviewable by EU courts.[16]

II. The factual and legal context

In 2008, the EU established a civilian “mission” called EULEX Kosovo. Essentially, the mission consists of a group of fonctionnaires tasked by the EU to support selected rule of law institutions in Kosovo (the judiciary, the police). Among its other tasks, EULEX is entrusted with investigating crimes allegedly committed during the war in Kosovo.[17] The mission was established under the EU’s Common Security and Defence Policy (CSDP), which is part of CFSP. The applicants, KS and KD, lost family members in the conflict, and brought an action for damages against the European Union (against the Council, the European Commission, and the European External Action Service) before the General Court, alleging, in essence, a breach of their fundamental rights because those crimes were not properly investigated.[18] The rights allegedly violated were arts 2, 3, 6, 13 and 14 ECHR (the right to life, the prohibition of torture, the right to a fair trial, the right to an effective remedy, and the right to be free from discrimination).

The General Court dismissed the action on procedural grounds finding it lacked jurisdiction to hear the case.[19] It reached that conclusion because arts 24(1) TEU[20] and 275 TFEU[21] limit the Court’s jurisdiction in the area of CFSP to cases concerning matters (institutional decision-making procedures and restrictive measures) that clearly do not relate to the situation of the applicants.[22] In addition, precedents could be distinguished because the facts giving rise to those judgments (and therefore the relevant provisions) concerned EU personnel or the EU budget.[23]

KS and KD appealed that order, arguing that the General Court misinterpreted those provisions as well as the case law, and asking the ECJ to set aside the order under appeal. The ambiguity over the extent of the Court’s jurisdiction stems from the fact that on several occasions the ECJ found that it could in fact hear claims in situations other than those mentioned in arts 24(1) TEU and 275 TFEU.[24] In particular, the applicants relied on Bank Refah, in which the Court held that it has jurisdiction over action for the damages caused by the wrongful inclusion of an applicant in restrictive measures of the EU.[25]

The main question before the European Court of Justice was therefore whether it has jurisdiction to hear actions for damages allegedly caused by the EU due to breaches of fundamental rights committed in the context of a CSDP mission.

III. The judgment of the European Court of Justice

The Court found in favour of KS and KD, partially setting aside the order under appeal.

The ECJ grouped the arguments of the appellants into two clusters, those relating to the relevant Treaty provisions on CFSP (arts 24 TEU and 275 TFEU), and those relating to the case law on personnel management in the context of CSDP missions.

On the first cluster, the Court acknowledged that the acts and omissions referred to by KS and KD did not concern the monitoring of compliance with art. 40 TEU or the review of such individual restrictive measures.[26] The Court then made two main arguments to explain why the General Court was not wrong in finding that it did not have jurisdiction based on an interpretation of arts 24 TEU and 275 TFEU.[27] The first was that the principle of conferral and for the balance of institutional power (to which arts 24 TEU and 275 TFEU give expression) could not be upset by the mere fact that a violation of fundamental rights was alleged.[28] Otherwise, those articles “would be deprived of their effectiveness in part and the principles of conferral and of institutional balance infringed”.[29] By contrast, the Advocate General had suggested that compliance with fundamental rights must be guaranteed for any act or omission of the EU, regardless of its legal basis. She had concluded that “the EU Courts’ jurisdiction to review any CFSP measure, including a political or strategic one, in order to ensure its conformity with fundamental rights cannot be excluded by Article 24(1) TEU and Article 275 TFEU”.[30] This was because, in the Advocate General’s reasoning, the value of the rule of law expressed in art. 2 TEU “requires that both EU and Member State authorities be subject to judicial review”.[31] This requirements also applies to CFSP, because that policy is part of the EU legal order (arts 21 and 23 TEU).[32]

The second argument of the ECJ relates to ensuring an equivalent level of protection between the right to an effective remedy of art. 47 Charter, on the one hand, and the right to a fair trial and to an effective remedy in arts 6(1) and 13 ECHR, on the other hand.[33] Those are not absolute rights and may be restricted.[34] In particular, a legitimate restriction is “the institutional balance between the executive and the national courts; that institutional balance may be reflected in a constitutional limitation of the jurisdiction of the courts of a State as regards acts that cannot be detached from the conduct by that State of its international relations”.[35] In addition, the provision that the EU shall accede to the ECHR (art. 6(2) TEU) does not extend to the jurisdiction of the Court of Justice of the European Union in relation to the CFSP. This is because the first sentence of art. 2 of Protocol 8 of the Lisbon Treaty provides that “the agreement [relating to that accession] shall ensure that accession of the Union shall not affect the competences of the Union or the powers of its institutions”.[36] Further, the judgment in Bank Refah cannot be extrapolated to the present situation, because that case concerned restrictive measures and “it is the individual nature of those measures which, in accordance with the second paragraph of Article 275 TFEU, permits access to the Courts of the European Union”.[37] The fact that the Court has exclusive jurisdiction over actions based on the non-contractual liability of the EU is not a reason to conceive of that jurisdiction to be unlimited and extending also to any CFSP acts.[38]

Turning then to the second cluster, that is the case law on personnel management in the context of CSDP missions, the Court outlined the criteria for when it has jurisdiction in situations that are outside those expressly provided for in arts 24 TEU and 275 TFEU. In those cases, “the jurisdiction of the Court of Justice of the European Union may be based on the fact that the acts and omissions at issue are not directly related to the political or strategic choices made by the institutions, bodies, offices and agencies of the Union in the context of the CFSP, and in particular the CSDP”.[39]

In the light of this finding, EU courts have jurisdiction over the pleas by KS and KD that EULEX Kosovo allegedly lacked appropriate personnel because, the Court found, this is not directly linked to political or strategic choices;[40] the provision of legal aid in proceedings before the review panel of EULEX Kosovo;[41] the powers conferred to the review panel or of remedies for breaches;[42] and the failure to remedy the breaches of fundamental rights found by the review panel.[43] However, it does not have jurisdiction over the plea that EULEX Kosovo allegedly lacked appropriate resources (because “the resources made available to a CFSP mission, and in particular a CSDP mission, on the basis of the first subparagraph of Article 28(1) TEU, are directly related to the political or strategic choices made within the framework of the CFSP”).[44]

Since it had jurisdiction, but it did not have sufficient elements to pass judgment on the merits, the ECJ referred the case back to the General Court.

IV. Analysis

The judgment in KS and KD raises several legal issues, and perhaps also questions of judicial politics. This annotation discusses the reasoning of the Court, showing how it is in line with precedent and how it strikes a fair balance between the principle of conferral and institutional balance on the one hand, and the principle of effective judicial protection on the other hand; it also discusses some implications of that reasoning for future cases that may arise in the field of CFSP. Other important aspects, including the relationship between the EU and the ECHR after KS and KD, are the object of other contributions.[45]

iv.1.  Acrobatic but Defensible Reasoning: the role of Precedent and of the Treaties

KS and KD shows how the ECJ deals with precedent in its recent case law. This is what its President refers to as the “stone-by-stone” approach,[46] which builds, incrementally, the lines of its case law starting from a general and somewhat vague statement contained in one (seminal) judgment, and then further refining it (or attaching precise meaning to it) in subsequent cases, as needed according to the accidents of litigation. Some of these foundational formulas are the “genuine enjoyment of the substance of rights” in Ruiz Zambrano,[47] or the “status of citizen of the European Union is destined to be the fundamental status of nationals of all the Member States” in Grzelczyk,[48] and so on.

For KS and KD, the “foundational stone” is the statement in Opinion 2/13 that “as EU law now stands, certain acts adopted in the context of the CFSP fall outside the ambit of judicial review by the Court of Justice”.[49] The question of what exactly is within judicial review (and therefore what is not) is still relevant because the EU shall accede to the ECHR. In Opinion 2/13, the Court said that it could not. In KS and KD, the Court implied that it still cannot. In Opinion 2/13, the ECJ had been asked whether a draft agreement on the accession of the EU to the European Convention for Human Rights was compatible with EU law. The Court found that it was not, inter alia because the agreement would have entailed that the ECtHR would have had jurisdiction over all CFSP acts, whereas the ECJ does not have jurisdiction on at least some of them. The Court also found that such jurisdiction of the ECtHR would be intolerable in the Union’s legal order because “jurisdiction to carry out a judicial review of acts, actions or omissions on the part of the EU, including in the light of fundamental rights, cannot be conferred exclusively on an international court which is outside the institutional and judicial framework of the EU”.[50] Why could such jurisdiction not be conferred exclusively to a judicial body outside the EU legal system? The reasoning is not entirely clear from the quoted passage of Opinion 2/13. In my view, the answer is that the absence of the Court’s power to carry out judicial review of some foreign policy decisions is a special characteristic of the EU legal order. This is what the parties (the EU Member States) intended when drafting the Lisbon Treaty: some foreign policy acts should not be judicially reviewable.[51] It follows that the EU cannot accede to the ECHR unless either the accession clearly makes an exception for these foreign policy decisions, and/or the EU Treaties are changed.

The case law supports such an argument. When discussing the lack of jurisdiction over CFSP, in the abovementioned paragraph of Opinion 2/13 the Court refers to Opinion 1/09 (on the Patent Convention Agreement) as an authority. That Agreement incorporated EU law provisions into the substantive law that was to be applied between the parties, and, in particular, by a European Patent Convention Court envisaged by such agreement. In Opinion 1/09, the rationale for the finding that the EU cannot confer jurisdiction exclusively to a body outside the EU judicial order was that the jurisdiction of an international court (there, the European Patent Convention court) “would deprive courts of Member States of their powers in relation to the interpretation and application of European Union law and the Court of its powers to reply, by preliminary ruling, to questions referred by those courts and, consequently, would alter the essential character of the powers which the Treaties confer on the institutions of the European Union and on the Member States and which are indispensable to the preservation of the very nature of European Union law”.[52] If the ECtHR were to have jurisdiction over CFSP acts, what is the “essential character of the powers” conferred by the Treaties that would be altered? The answer may be found in Opinion 2/13 itself, which refers to the exclusion of the Court’s jurisdiction in CFSP as a “situation” that is “inherent to the way in which the Court’s powers are structured by the Treaties, and, as such, can only be explained by reference to EU law alone”.[53] KS and KD is perfectly in line with that reasoning: the exclusion of the Court’s jurisdiction from CFSP is a reflection of the principles of conferral and of institutional balance.[54] This situation, the limited jurisdiction of the Court, is, as Opinion 2/13 clarifies, “a specific characteristic of EU law”.[55] If one were to leave aside KS and KD for a moment, and ask why such an essential characteristic cannot be affected, i.e. why the autonomy of EU law must be protected even in this case, the implicit rationale of Opinion 2/13 arguably flows from the political bargain struck between the Member States. It is an instance of the principle nemo dat quod non habet (no one can give what they do not have): since the Member States decided that the EU Courts cannot review certain (political) choices, the EU itself cannot delegate that power (because it does not have it) to Strasbourg (or to any other international court outside the EU legal system). [56]

But this is not the only line of cases that flows from Opinion 2/13. In later cases, little by little, or stone-by-stone, the boundaries of the jurisdiction of the Court in CFSP acquired more precise contours. Thus in Rosneft the Court confirmed that in the context of sanctions it has jurisdiction only over decisions providing for restrictive measures against natural or legal persons (“individual decisions”),[57] and not the provisions of “general application, in that they impose on a category of addressees determined in a general and abstract manner a prohibition on making available funds and economic resources to entities listed in their annexes”;[58] in Elitaliana[59] and in H,[60] the Court held that acts related to budgetary measures or of staff management respectively are subject to judicial review, even though they are contained in CFSP decisions; in Bank Refah, the Court found that also action for damages stemming from the wrongful inclusion in EU restrictive measures against natural or legal persons were within the Court’s jurisdiction. Now, KS and KD confirms that the jurisdiction covers also actions for damages stemming from other actions or omissions of the EU under CFSP. And even though the Court refers to the precedents of Elitaliana, H, and SatCen for the idea that it will not have jurisdiction on acts or omissions “directly related to strategic and political choices”,[61] such a notion is not to be found in those precedents, at least not identified with these words nor in such an explicit way. Those cases merely hold that the Court does have jurisdiction on acts relating to staff management or to public procurement. Admittedly, some re-interpretation of the wording of the case law – which was in itself an interpretation of the Treaties – is necessary, but KS and KD is also in line with Elitaliana, H, and SatCen. To be clear: those cases told us some instances in which the Court does have jurisdiction over CFSP acts; KS and KD now re-reads those case as also containing the general rule for when the Court does not have jurisdiction.

Before turning to the crux of the matter, that is, what exactly is excluded from the Court’s jurisdiction (what “relates directly to a political or strategic choice”), few words may be said about the reasoning of the Court. As the previous paragraphs of this section shows, and as KS and KD clearly explains, there are two strands of cases on CFSP jurisdiction. One could be termed the “limits” case law, which starts with Opinion 2/13 and of which KS and KD is an illustration: the wording of arts 24 TEU and 275 TFEU “would be deprived of their effectiveness in part and the principles of conferral and of institutional balance infringed”[62] if it was simply ignored. Art. 47 Charter and art. 6 and 13 ECHR cannot call this assessment into question (in essence, I read the Court as saying here that the AG went a step too far in her Opinion in KS and KD).[63] In these cases, the Court interprets the limits broadly, in the name of the effectiveness of arts 24 TEU and 275 TFEU, and in light of the principles of conferral and institutional balance (of which they are an expression).

The other line of cases includes Mauritius,[64] Elitaliana, H. Here, it is instead the Court’s jurisdiction that is interpreted broadly (some could say overly stretched) in the light of general principles, namely judicial protection as manifestation of the rule of law.[65] In both strands of cases, the Court could therefore reasonably claim that it has neither “expanded”[66] nor “limited”[67] its jurisdiction on CFSP but has simply “found” its correct extent by interpreting the letter of the law in the light of a hierarchically superior principle.[68] The issue is construed by the Court as a “clash” between general principles:[69] on the one hand there is the principle of conferral and of institutional balance; on the other, the rule of law and judicial protection. As far as CFSP is concerned, these are given concrete expression in arts 24(1) and 275 TFEU, and in art. 47 Charter on the other hand. They are not easy to square, and the balance is drawn by the Court in ways that raised objections.[70] In my view, the position of the Court is defensible (although the reasoning whereby the conclusion is reached is not immune from criticism), for the reasons mentioned in the introduction. I try to offer a full account for these statements elsewhere.[71]

iv.2. What is a political or strategic choice?

In KS and KD, the Court gave a vague answer to the repeated question of what the boundaries of its jurisdiction over CFSP provisions and acts are (it shall be recalled that the Advocate General suggested that the Court has jurisdiction to review any act on the grounds of fundamental rights).[72] The test is now as follows: in the context of CFSP, the Court does not have jurisdiction on “acts or omissions directly related to […] political or strategic choices [of the EU]”.[73] This is not found in the Treaties, at least not expressis verbis. We now know from KS and KD that, implicitly, this is what arts 24 TEU and 275 TFEU mean.

There are some tentative indications of what may amount to a political or strategic choice of the Union in KS and KD,[74] but even fewer as to when an act or omission relates “directly” to these. The usual criteria of interpretation would suggest that we should start by looking at the wording of the Treaties: art. 26(1) TEU states that “The European Council shall identify the Union's strategic interests”, on the basis of which the Council shall take decisions necessary to define or implement them (art. 26(2) TEU); the Political and Security Committee (PSC) may be authorised by the Council “to take the relevant decisions concerning the political control and strategic direction of [a crisis management] operation” (art. 38 TEU). All these (and those by the Council authorising the decisions by the PSC) may be regarded as not subject to judicial review. There hardly seems to be a meaningful distinction between “strategic” and “political”, in this context. Looking at the case law, some of the elements that the Court is likely to take into account when determining what is “political or strategic” are whether an act affects the rights of individuals in any immediate way (if it does affect them, it should not be simply regarded a strategic or political choice) or whether it concerns budget or personnel (in which case it seems to be operational – at least this is how the Court’s case law has constructed this category). Some situations are clearer than others: restrictive measures which are not “against natural or legal persons” but are of general application are not subject to judicial review. Thus, the decision to sanction the Russian gas sector, imposing a requirement of prior authorisation on all EU operators who want to transfer equipment to companies operating in that sector (this what challenged in the litigation that resulted in the judgment in Rosneft[75]), is of general character and not subject to judicial review.[76] What is “political or strategic” about it is that it is the result of a sovereign choice of the Union that the EU Treaties have not made subject to any other condition than “the international situation so requires”,[77] or to no condition at all,[78] thus giving no guidance to the institutions as to when something needs to be done, or as to what do to when something needs to be done.[79] In the recent practice of the EU, an example of that choice is the decision to sanction Russia as opposed to Ukraine, or, more realistically, to sanction the Russian gas sector – at a significant cost to EU citizens – as opposed to or in addition to sanctioning, say, the Russian tourist industry. Further, a political or strategic choice is the one relating to the resources given to a CSDP mission.[80] Another (this time, speculative) example is the decision of the EU to deploy a mission.[81] There are situations in which the very decision to deploy or not to deploy a mission, or to sanction or not to sanction a country might violate fundamental rights or other rules of international law binding upon the EU. For example, the EU could violate (by action or inaction) the right to self-determination of people in a certain part of the world, or (by inaction) other fundamental rights if a duty of humanitarian intervention under international law exists.[82] However, a plausible reading of KS and KD is that the fundamental Treaties do not prohibit this violation to occur if they subtract such choices from judicial review.

When does an act or omission relate “directly” to a strategic or political choice? Since there is no guidance on this notion in the judgment under analysis, it is probably best understood as a “hedging” strategy, a test through which the Court adds further ambiguity as to what a precise answer will look like. One could envisage a distinction between a decision that relates “directly” (for example, because it allows or forbids or defines or implements a strategic/political decision without further intermediate steps[83]) and one that relates only “indirectly” (for example, because it follows, as a second step, from a decision that related directly). Contentious or borderline cases are bound to arise regardless of the precise formulation, and the twin notions of “directly related to” and of “political or strategic choice” give sufficient leeway for the Court to define the boundary on a case-by-case basis as necessitated by the factual and legal circumstances arising out of litigation.

In fact, a CFSP act or omission directly relates to political or strategic choice whenever the Court will hold that it directly relates to a political or strategic choice. It is likely that the Court will be asked to clarify what this means in the future and may possibly need to give a general answer that is not circumscribed to the specific circumstances of the case. A general answer may be needed, for example, in a preliminary ruling procedure, where the ECJ needs to provide guidance on the interpretation of the Treaties in a way that is not dependent on the facts of the main proceedings (KS and KD is a direct action, in which the Court’s judgment, and especially the judgment of the General Court, cannot do away with the facts of the litigation).

V.  Conclusion

The European Court of Justice’s judgment in KS and KD is an important and welcome development in the interpretation of EU jurisdiction over actions related to the Common Foreign and Security Policy. The Court essentially established jurisdiction to hear action for damages stemming from acts or omissions committed by the EU in the context of CSDP activities (decisions, omissions, civilian missions, military operations, and so on). The only exceptions are those of a political or strategic nature. In affirming its jurisdiction, the Court opens a path for individuals affected by EU missions to seek judicial remedies for violations of their rights. The ruling shows the difficulty in balancing the principle of conferral (which limits the Court’s jurisdiction in CFSP) with that of effective judicial review (which requires an effective remedy for any violation of rights). The Court tried to do so by stating that political and strategic decisions remain outside judicial scrutiny. What this means in practice will need to be decided in future cases. The judgment is admittedly in tension with the wording of the Treaties, because the outcome is determined by a systematic and teleological interpretation. The reasoning is, however, fully in line with over a decade of case law, as this annotation has sought to demonstrate.