Article 63 as Evidence Law: The Gambia v Myanmar and the Missing Intervener

[Khan Khalid Adnan serves as the Head of the Chamber at Khan Saifur Rahman & Associates, Dhaka, Bangladesh. He is a Fellow of the Chartered Institute of Arbitrators (FCIArb), a Barrister in England and Wales, and an Advocate of the Supreme Court of Bangladesh.]

In the genocide docket, Article 63 interventions under the ICJ’s Statute are no longer a neutral “interpretation-only” device; in The Gambia v Myanmar they have become a form of collective evidentiary calibration—especially on genocidal intent and standards of proof—binding interveners and shaping the merits ex ante. That is precisely why Bangladesh, despite its Article IX reservation to the Genocide Convention 1948, should enter now: either through a merits-phase Article 63 limited to Articles I–VI, or via Article 62 to protect its concrete legal interests.

From Interpretive Right to Evidentiary Rule-making

On 25 July 2025, the ICJ admitted four additional Article 63 declarations in The Gambia v Myanmar—Slovenia, the DRC, Belgium and Ireland—after the Court’s first wave in July 2024 (Maldives; and a joint declaration by Canada, Denmark, France, Germany, the Netherlands and the UK). The 2025 Order reiterates the familiar formula: intervention under Article 63 is limited to the construction of the Genocide Convention; parts that go to facts or application will be disregarded; other rules of international law may be considered only in so far as they assist treaty construction (VCLT, Art. 31(3)(c)). It also applies the amended Rule 82 of the Rules of Court (in force from 1 June 2024), holding that the “as soon as possible” requirement is read together with the “no later than the last written pleading” deadline and was met in this case. The 2025 Order itself crystallises the point: the declarations may speak to Articles I–VI (including the definition of dolus specialis in Article II) because that construction is in issue at the merits; anything beyond construction will simply not be considered (paras 58-60). 

What makes this moment different is Judge Cleveland’s declaration appended to the 2025 Order. She states that the standard of proof for genocidal intent is part and parcel of the construction of Article II, because “gravity” and evidentiary thresholds flow from the rights and obligations the Convention creates (paras 5-6). That moves questions like the inferential threshold for dolus specialis out of the realm of free-floating procedure and squarely into treaty interpretation—the very domain of Article 63. 

Once evidence law is re-described as treaty construction, Article 63 interventions are not just amici illumination. They become a binding interpretive caucus on evidentiary baselines: by operation of Article 63(2), the construction ultimately given by the Court will be “equally binding” on interveners. That is normatively attractive for coherence across parties to the Convention; but it also affects the merits, because it sets the yardstick by which genocidal intent will be inferred from patterns of conduct and official statements, as a matter of evidence (Judge Cleveland, para 4, fn 5).

If the 2025 Order is read alongside Judge Cleveland’s declaration, its conclusions diverge significantly from those in the Ukraine v Russia Order of 5 June 2023, where the notion of specific intent was not even addressed (para 84), and from the 2024 The Gambia v Myanmar Order, in which the Court emphasised the need to interpret Article II of the Genocide Convention defining specific intent (para 43). By contrast, the 2025 Order explicitly stated that the definition of specific intent in Article II is central to construing the Convention (para 59), with Judge Cleveland going further by highlighting the long-standing disagreement between the parties over what constitutes adequate evidence of dolus specialis, including the applicable standard of proof (para 4). This marks a considerable expansion of the Court’s approach, moving well beyond what Judge Ronny Abraham, the former President of the Court, once envisaged when he described the object of Article 63 intervention as being very limited.

The Comparative Picture: Ukraine v Russia and South Africa v Israel

The mass-intervention template is Ukraine v Russia. On 5 June 2023, the Court admitted a historically large set of declarations under Article 63 while excluding the United States because its reservation meant that the judgment’s construction of Article IX could not be “equally binding” upon it—confirming the provision-by-provision logic noted above. The result was an unprecedented collective interpretation exercise of the Genocide Convention’s core articles. 

That scale raised equality-of-arms concerns long flagged in earlier cases. In Whaling in the Antarctic, the Court admitted New Zealand’s Article 63 intervention but noted that the intervener is not a party; Judge Owada cautioned that coordinated interventions can look like “active collaboration in litigation strategy” (para 5) and so risk fairness if unmanaged. 

The Court’s 2024 amendments to Rule 81–82 of its Rules are the institutional answer so far. They front-load timing (Article 63 declarations must now be filed no later than the last written pleading) and codify the Court’s ability to manage duplicative filings. The 2025 Order in The Gambia v Myanmar applies those changes and, crucially, draws a line around evidentiary material: interveners may discuss evidentiary standards as part of construction, but cannot introduce or weigh case-specific evidence. That is precisely how Article 63 becomes evidence law by another name without letting interveners litigate the facts.

In South Africa v Israel, we see the same dynamics, but on a rolling timetable. Ireland filed an Article 63 declaration on 6 January 2025; Brazil announced on 23 July 2025 that it would intervene. Meanwhile, by Order of 14 April 2025, the Court extended Israel’s Counter-Memorial deadline to 12 January 2026—a schedule that will have to accommodate multiple interveners’ written observations. At least 13 States have by far filed their declarations of intervention. The cases of The Gambia v Myanmar, Ukraine v Russia, and South Africa v Israel clearly illustrate how large-scale interventions have now become a common feature before the Court—a notable shift given that, until October 2017, only four Article 63 intervention declarations had been made since the Court’s establishment.

Why Bangladesh’s Silence is Legally Costly—and Legally Unnecessary

Bangladesh hosts the largest population of Rohingya refugees and is not a bystander to the legal consequences of any final judgment in The Gambia v Myanmar. While finding jurisdiction, the Court itself acknowledged this (Judgment on Preliminary Objections, para 113). Yet Dhaka has not intervened. The reason most commonly given is Bangladesh’s reservation to Article IX of the Genocide Convention—“consent of all parties” to the Court’s jurisdiction—assumed to bar an Article 63 intervention. The reservation is real. 

But that is not the end of the analysis. In Ukraine v Russia, the Court held that a State cannot intervene as to the construction of a provision to which it has reserved—hence the United States’ inadmissibility at that stage—but did not foreclose interventions as to other provisions of the Convention by which the State is bound. The 2025 Order in The Gambia v Myanmar makes the same structural point from the other end: Articles I–VI (including the Article II definition and intent standard) are squarely “in question” at the merits, and that is what Article 63 interventions may address. Nothing in Bangladesh’s reservation excludes intervention on those provisions. 

If Dhaka remains unconvinced about Article 63, Article 62 is still open. Intervention by a non-party is permitted where a State shows an “interest of a legal nature which may be affected” by the decision; that test is preventive and does not convert the case into a trilateral dispute. Classic examples include Equatorial Guinea in Cameroon v Nigeria (maritime delimitation) (paras 2, 5, 12) and Greece in Jurisdictional Immunities (paras 5-6, 22, 25-29, 31). Bangladesh can articulate at least three cognizable legal interests: the conditions for safe and dignified return; the scope of cooperation duties (including evidence preservation and mutual legal assistance) on its territory; and the calibration of inferential standards for dolus specialis that will govern future accountability dialogues. That framing tracks the Court’s Article 62 practice and the intervention’s preventive function. 

The deeper worry in Dhaka may be binding effect and litigation strategy: under Article 63(2) a merits-phase intervention binds Bangladesh to the Court’s construction; under Article 62 it does not (Article 59 still governs res judicata). But that is a feature, not a bug, if the issue at stake is the very standard of proof that will structure Rohingya repatriation-related obligations and any future inter-State accountability. Sitting out leaves those evidentiary baselines to others. 

Guardrails—and a Concrete Path for Bangladesh

If Article 63 is now an evidentiary calibration tool, two guardrails are essential. First, the Court should sequence and consolidate interventions to prevent a readability gap between one respondent and dozens of coordinated interpretive positions. The 2024 amendments empower this; the 2025 Order already instructs interveners to stay within construction and foreshadows tighter control at oral phase (para 60, Rule 86). Second, the Court should encourage issue-lists that focus precisely on Article II intent and the standard of proof, areas where third-State views advance consistency without re-litigating facts—mirroring Judge Cleveland’s approach. 

For Bangladesh, the legally safest route is a two-step filing. Step one: an Article 62 application identifying its legal interests in the conditions for return, co-operation duties, and evidentiary handling of material generated on its territory, emphasising that Article 62 is preventive and does not seek party status. The jurisprudence in Cameroon v Nigeria (Equatorial Guinea) and Jurisdictional Immunities (Greece) shows the Court has granted such targeted interventions, particularly where neither party objects. Step two: if Dhaka wishes to help shape the construction of Articles I–VI, a merits-phase Article 63 declaration limited to those provisions would avoid the Article IX barrier while engaging the very standard-of-proof questions that will govern the Rohingya file for years. The Court’s 2025 Order makes explicit that such interventions are admissible; it is better to influence the yardstick now than to live with one set by others. 

Bangladesh might worry about politicisation or being “bound” to a construction that later constrains its diplomacy. But recall why Article 63 binds interveners: to stabilise treaty meaning across parties. In a case where Bangladesh’s day-to-day responsibilities are already shaped by how other States and tribunals understand the Convention, participating to calibrate the intent-inference and proof rules is rational risk management, not over-reach. The alternative—silence while a growing group of States converge on a doctrinal package—creates precisely the interpretive coalition Bangladesh should want to shape. 

Conclusion

The novelty in The Gambia v Myanmar is not that more States intervened; it is that the Court, and a member of the Bench, have folded standards of proof and intent-inference into “construction” under the Genocide Convention. Article 63 has always been about interpretation; in genocide litigation it now settles evidence law too. Article 63, once largely symbolic with little practical significance, has recently gained greater relevance in interpreting the convention at issue to safeguard collective interests, reflecting the extensive collaborative effort among intervening States and highlighting concerns over “equality of arms.” As illustrated in The Gambia v Myanmar, interventions now have the potential to serve as a mechanism for collective evidentiary calibration, marking a profound evolution in the jurisprudence of Article 63. That is a healthy development so long as the Court uses its amended Rules to manage mass participation and to keep interveners on the legal (not factual) plane. For Bangladesh, the legal road is open—Article 62 to protect immediate interests, and, if it wishes, a targeted Article 63 to help set the evidentiary yardsticks that will bind everyone else anyway. Waiting only cedes the field.