Red lines in a grey zone: the validity of a Russia-Ukraine peace treaty under international law

While peace still appears remote, the prospect of a negotiated end to the war in Ukraine raises difficult questions concerning the international law governing inter-state peace treaties. The unlawfulness of Russia’s use of force and the peace plans proposed to date point to issues of validity, particularly under Articles 46, 52 and 53 of the Vienna Convention on the Law of Treaties (VCLT).

This post discusses the extent to which the international law of treaty validity constrains the permissible terms and circumstances of concluding a peace treaty. It demonstrates that, in the context of a potential Russia–Ukraine peace treaty, international law requires respect for Ukraine’s fundamental constitutional constraints, safeguards its freedom to consent to a treaty, and precludes treaty terms that conflict with peremptory norms. Yet, construed along the lines advanced in this post, these constraints leave significant room for accommodating the imperatives of bringing a militarily unresolved conflict to an end.

Validity in light of Ukraine’s constitutional constraints on treaty-making

A potential peace treaty between Russia and Ukraine could raise questions under Article 46 VCLT where treaty terms require territorial change or other constitutional amendments but the relevant constitutional constraints are not observed. Article 46 provides a narrow ground for invalidating a State’s consent to be bound by a treaty, where that consent was expressed in manifest violation of an internal rule of fundamental importance governing competence to conclude treaties.

According to Article 9 of the Constitution of Ukraine, international treaties contrary to the Constitution may be concluded only after the relevant constitutional amendments have been adopted. This requires consideration of three constitutional amendment procedures potentially relevant to a peace treaty.

First, Article 157 of the Constitution prohibits amendments that abolish or restrict human rights, undermine independence, or violate the territorial integrity of Ukraine. Instead, alteration of the territory of Ukraine, as defined in Article 133, requires an All-Ukrainian referendum designated by the parliament of Ukraine (Article 73). Ukraine’s legal framework on referenda contains a seeming contradiction: territorial changes must be decided by referendum, yet issues affecting territorial integrity cannot be submitted to one (Art. 3(2)(3), Law 1135-IX). However, the law explicitly requires that any international treaty involving territorial change be approved by a binding referendum. Read together with Article 73 of the Constitution, and in light of President Zelensky’s statements, this suggests that a referendum on territorial change must be regarded as legally conceivable. While a referendum is clearly required for de jure territorial change, de facto territorial arrangements or establishment of a special economic zone may not require a referendum.

Second, Article 156 of the Constitution establishes a heightened amendment procedure for Chapters I, III, and XIII of the Constitution, which is potentially relevant where a peace treaty affects foundational constitutional principles, such as those concerning official language. Other constitutional amendments are governed by the ordinary procedure under Article 155, and may be required, for example, where a peace treaty addresses Ukraine’s future membership in the European Union or the North Atlantic Treaty Organization given the references in Articles 85, 102 and 106 of the Constitution. Finally, constitutional amendments are prohibited altogether during martial law or a state of emergency (Article 157(2)), and an opinion by the Constitutional Court on any constitutional amendment is required (Article 159).

The ground of invalidity recognised in Article 46 is exceptional in character and interpreted narrowly. Although some commentators confine it to procedural rules governing treaty-making authority (Corten et al (eds), VCLT Commentary, p. 1239), certain substantive constitutional limitations can also be understood as falling within its scope (Dörr/Schmalenbach (eds), VCLT Commentary, p. 849; YILC 1966, vol II, p. 240). While practice under Article 46 is limited, Article 9 of the Ukrainian Constitution can be regarded as governing treaty-making competence in a broad sense.

Although violations of internal law will rarely meet the strict threshold of Article 46, a peace treaty concluded in disregard of Article 9 and the associated amendment procedures could arguably fall within this exception. First, the constitutional provisions concerned are of clear “fundamental importance.” Second, the requirement of a “manifest” violation may also be satisfied, as one “objectively evident to any State conducting itself in accordance with normal practice and good faith” (Article 46(2) VCLT). While this is ultimately a context-specific assessment (YILC 1966, vol II, p. 242), indicative factors include whether requirements were “easily ascertainable and a matter of common knowledge” (Rosenne, p. 207) and “properly publicized” (ICJ, Boundary between Cameroon and Nigeria [2002] ICJ Rep 303, para 265). In the case of Ukraine, constitutional restrictions relating to treaty ratification and constitutional amendment are objectively evident, publicly available, and have been extensively discussed internationally since 2022.

Accordingly, unless Ukraine were to amend or replace its Constitution in advance of a peace treaty to remove these constitutional requirements on treaty-making and constitutional amendment, and suspend martial law to permit such amendments, a treaty concluded contrary to these constitutional requirements may lead to an invocation of Article 46 VCLT.

The practical reach of this ground for invalidity is limited as invalidity under Article 46 is not ab initio and must be invoked by the State in question. That said, discussions in Ukraine (see, e.g., here and here) suggest that such invocation may be possible. However, if Ukraine subsequently accepted the treaty’s validity or acquiesced through its conduct, the right to invoke this ground for invalidity would be lost (Article 45, VCLT). A future government would then also be unable to reopen the issue.

Peacemaking in coercive circumstances

Under Article 52 VCLT, a treaty is void ab initio if its conclusion has been “procured by” an unlawful threat or use of force. As the unlawfulness of Russia’s use of force is established, the central issue is under which circumstances a peace treaty intended to bring the conflict to an end would be “procured by” the aggression which would thus vitiate Ukraine’s consent to be bound.

In the context of peace treaties, Article 52 remains a Sword of Damocles that has never fallen (Forlati, p. 322). In brief, post-VCLT practice demonstrates that (i) few peace treaties were concluded following unlawful threat or use of force in this period; (ii) although not directly comparable to the Russia–Ukraine context, some of these treaties included concessions by the victim State; and (iii) Article 52 was nevertheless not invoked to invalidate such agreements.

The key interpretive divide under Article 52 concerns the meaning of a treaty being “procured by” coercion and, specifically, the causal link required between unlawful force and treaty conclusion. An expansive conception of causality suggests that coercion leads to invalidity where it significantly influenced the conclusion of a treaty. A stricter interpretation instead would require direct causality: a treaty would be invalid only where the affected State was “reduced to such a degree of impotence as to be unable to resist the pressure to become a party to a treaty” (Lauterpacht, para 7). The latter view appears more consistent with the limited existing practice (Corten et al (eds), VCLT Commentary, p. 1379).

Some of the divergence of opinion in the Russia-Ukraine context can be traced to the distinction between the expansive and strict approaches to causality. For some commentators, as long as Russia’s aggression remains a significant factor in securing Ukraine’s consent to concessions, any resulting agreement would be invalid under Article 52 (Fox; Hollis; Vishchyk/Pizzi). Others maintain that, so long as Ukraine retains a choice whether to conclude the treaty, concessions, including, for a narrower group, territorial concessions, would not necessarily render the agreement invalid (Herbert; Bella). In such circumstances, the treaty would render indirect the causal link between the concession and the unlawful use of force.

Article 52 does not preclude the validity of peace treaties involving concessions of any kind, provided that the victim State has a meaningful role in determining the terms and retains a choice whether to consent, even where that choice is undoubtedly influenced by coercive circumstances. Two considerations support this interpretation. First, Article 52 is structurally concerned with a prohibited method of treaty conclusion, i.e., the conditions under which consent is produced, rather than prescribing or prohibiting specific substantive outcomes. It is therefore also difficult to distinguish categorically between different forms of concession for the purposes of Article 52, and to suggest that the provision effectively prohibits territorial concessions while potentially permitting concessions of other kinds.

Second, although it is widely accepted that the substantive terms of an agreement may provide evidence regarding the surrounding coercive circumstances, alongside other circumstantial factors such as the degree of force used and whether a treaty is concluded in conditions of ceasefire, neither consideration alone can determine whether consent was procured by coercion. The argument that “State A would never have accepted the concession in question had it not been reduced to a position in which it could no longer freely consent” infers coercive circumstances solely from the terms of a treaty and risks circularity.

The diversity of views in the Russia-Ukraine context is rooted not only in differing legal interpretations, but also in differing assessments of the factual circumstances. One may conclude that any concession would invalidate a peace treaty, despite adopting a strict approach to causality, by considering that the facts on the ground do not leave the victim State any other option than to agree to a treaty (eg Kriener/Peters). On the other hand, it could also be argued that, since Ukraine has so far been able to refuse certain proposed settlement terms, it has not been “reduced to such a degree of impotence as to be unable to resist the pressure to become a party to a treaty” (to use Lauterpacht’s words). The validity of any future peace treaty under Article 52 will necessarily depend on the political and military circumstances prevailing at the time of its conclusion, including, for example, whether a treaty is concluded against the backdrop of a military stalemate rather than one side’s dominance, and whether the negotiation process is multilateral or held under the auspices of an international organisation.

Jus cogens invalidity of a peace treaty

Article 53 VCLT provides that a treaty is void “if, at the time of its conclusion, it conflicts with a peremptory norm of general international law”. This raises the question of what it means for a treaty to “conflict” with a peremptory norm.

In the context of a Russia-Ukraine peace treaty, some commentators adopt a broad interpretation of “conflict,” in arguing that any concession in a peace treaty would be in conflict with the peremptory prohibition of the use of force or, in case of territorial concession, with the right of self-determination or the prohibition of annexation, assuming their peremptory status (Fox; Prantl/Schmalenbach; Schmitt; Vidmar, p. 55). This interpretation is underpinned, among other considerations, by an understanding that a “conflict” in this context arises not only when a treaty directly contravenes or authorises a breach of a peremptory norm, but also when it undermines a peremptory norm in a broader sense (see Orakhelashvili, p. 138-9). Others consider this to be an overly expansive interpretation (eg Herbert; cf Kolb, 124ff). Under a stricter understanding of “conflict”, Article 53 renders a treaty invalid only when it involves a direct contravention of a peremptory norm. On this approach, which appears more plausible in the domain of treaties, concessions in a peace treaty would not necessarily conflict with the prohibition of the use of force, unless they require, permit or authorise unlawful force.

This is not to suggest that direct conflict between the terms of a peace treaty and a peremptory norm is generally inconceivable. For example, the terms of a peace treaty may conflict with the right of self-determination, which is widely recognised as a peremptory norm, at least in the context of colonial domination and foreign occupation. Such a conflict may render the treaty void under Article 53, unless the apparent incompatibility can be resolved through interpretation (see, e.g. on the Oslo Accords, Declaration of Judge Tladi, ICJ Palestine AO, para 35).

In relation to the fate of a peace treaty with contested validity under Articles 52 and 53, a further point of contention concerns the role of the United Nations Security Council (UNSC). It is generally accepted that the UNSC cannot derogate from peremptory norms (Conclusion 16, YILC 2022, vol II) and therefore cannot render legally binding the terms of a peace treaty that conflict with a peremptory norm, should it be established that there is such a conflict under Article 53. However, this possibility cannot be ruled out for a treaty which may be invalid only under Article 52, for example by a binding resolution incorporating the treaty’s terms or demanding compliance with them (Forlati, p. 329; Dörr/Schmalenbach (eds), VCLT Commentary, p. 892). This issue becomes more significant under broader interpretations of Articles 52 and 53 that expand the practical reach of invalidity, which may then prompt attempts to circumvent the legal consequence of invalidity through UNSC ‘validation’. The significance of this debate is more limited, however, on the interpretations proposed in this post, under which treaty invalidity would arise in narrower circumstances in the first place. UNSC endorsement may nevertheless be sought to lend political support to a valid peace treaty.

Navigating international law in peacemaking

If, generally in the law of treaties, “validity is a topic shrouded in mystery and practically not terribly relevant” (Klabbers, p. 545), this has been even more so in relation to peace treaties. This is reflected both in the divergence of commentary on the application of particular grounds for invalidity to peace treaties and in the limited State practice, as parties have generally been reluctant to invoke invalidity once a peace treaty has been concluded.

In practice, questions of treaty validity may also be sought to be sidestepped altogether. It has been a feature of peacemaking practice that parties seek to navigate tensions arising from international law without directly addressing them. First, parties may employ ambiguity. This may concern the substance of particular provisions or the legal character of the instrument. For example, some arrangements evade clear categorisation as an international treaty in the sense of the VCLT, as in the case of the 2015 Minsk “Package of Measures” or the 2020 Armenia–Azerbaijan–Russia declaration. Second, parties may rely on avoidance. Issues which may otherwise give rise to validity concerns may be excluded from the agreement, left for future negotiations, or referred to international dispute settlement. Third, parties may opt for de facto or provisional arrangements that avoid formally resolving contested legal questions. For example, the issue of territorial control may be addressed without recognition of sovereignty or transfer of title, or through temporary arrangements including ceasefire lines and demilitarised zones (see here for options for a possible Russia-Ukraine treaty).

These peacemaking strategies do not resolve underlying legal tensions and will not necessarily be available, or desirable, in every negotiation process. Nor do they guarantee legal effectiveness. They nevertheless illustrate how parties may seek to diminish the risk of subsequent claims of invalidity.

This blogpost forms part of the Mini-symposium on international legal issues concerning a peace treaty for Ukraine. The introduction to the symposium by Anne Peters, Gregory Fox, and Florian Kriener can be found here. The symposium builds on presentations and discussions during the Conference on International Law and Peace in Ukraine organized at the Max Planck Institute for Comparative Public Law and International from April 17-18, 2026 and supported by the German Foreign Office.