Gene M Burd
Pierson Ferdinand, Washington DC
gene.burd@pierferd.com

Introduction

The enforcement of arbitral awards against sovereign states in US courts increasingly turns on a deceptively simple but foundational question: who decides whether an arbitration agreement was formed for purposes of jurisdiction under the Foreign Sovereign Immunities Act (FSIA)? This question sits at the intersection of public international law, domestic jurisdictional doctrine and core principles of contract formation. In recent years, it has assumed heightened importance as sovereign respondents more aggressively challenge enforcement actions by reframing treaty-coverage disputes as defects in consent itself.

Recent decisions of the US Court of Appeals for the District of Columbia Circuit reflect a maturing and increasingly disciplined approach. While US courts insist on an independent judicial determination of sovereign consent as a jurisdictional prerequisite, they have simultaneously resisted efforts to transform disputes over the scope or application of an arbitration agreement into jurisdictional bars. The US District Court for the District of Columbia’s December 2024 decision in Stabil LLC v Russian Federation provides a particularly clear illustration of how this line is drawn in practice.^[1]

FSIA and the arbitration exception

The FSIA provides the exclusive basis for obtaining jurisdiction over a foreign state in US courts. Section 1604 establishes a presumption of immunity, while section 1605 enumerates limited exceptions. Among them, the arbitration exception in section 1605(a)(6) withdraws immunity in actions to confirm or enforce an arbitral award made pursuant to an agreement to arbitrate that is governed by a treaty in force for the US, such as the New York Convention.^[2]

Critically, FSIA immunity is jurisdictional. If no exception applies, the Court lacks subject-matter jurisdiction and must dismiss the case without reaching the merits. As a result, the existence of an arbitration agreement between the claimant and the foreign state is not merely an element of enforcement; it is a threshold jurisdictional fact. At the same time, Congress did not intend the FSIA to serve as a mechanism for relitigating arbitral jurisdiction or the merits of treaty claims under the guise of sovereign immunity.

Formation versus scope in DC Circuit jurisprudence

The DC Circuit plays an outsized role in FSIA jurisprudence because the District of Columbia serves as the default venue for actions against foreign states lacking a sufficient territorial nexus to another judicial district.^[3] The Court has consequently developed the most substantial body of appellate precedent on the arbitration exception, and its decisions frequently operate as de facto national guidance in sovereign immunity cases.

The DC Circuit has repeatedly emphasised the need to distinguish between disputes that negate the formation of an arbitration agreement and those that concern its scope or applicability. In Chevron Corp v Republic of Ecuador and Stileks LLC v Republic of Moldova, the Court held that, for FSIA purposes, it is sufficient that an arbitral tribunal purported to render an award pursuant to an arbitration agreement.^[4] Whether the tribunal was ultimately correct as a matter of treaty interpretation generally goes to the merits under the New York Convention, not to jurisdiction.

This framework was reinforced and extended in NextEra Energy Global Holdings BV v Kingdom of Spain, where the Court addressed Spain’s attempt to invoke EU law to negate its consent to arbitrate under the Energy Charter Treaty (ECT).^[5] The DC Circuit held that Spain’s argument – that the ECT’s arbitration clause was invalid as it applied to intra-EU disputes under the Achmea and Komstroy decisions of the Court of Justice of the European Union – went to the scope of the arbitration agreement rather than its existence. The Court emphasised that ‘for jurisdictional purposes, the FSIA’s arbitration exception requires that the arbitral tribunal ‘purported to make an award pursuant to the [treaty], not that it in fact did so’. NextEra thus confirmed that even arguments framed as challenging the validity of consent are properly treated as scope disputes subject to deferential merits review as long as they depend on treaty interpretation rather than the fundamental question of whether any offer to arbitrate was ever made.

The distinction between jurisdictional and merits analysis frameworks was crystallised in Hulley Enterprises Ltd v Russian Federation.^[6] There, the DC Circuit made clear that the existence of an arbitration agreement is a jurisdictional fact under the FSIA that must be determined independently and de novo by the Court. That holding fits squarely within the Court’s earlier decisions in Chevron and Stileks, which likewise recognised that FSIA jurisdiction turns on the existence of an agreement to arbitrate, while disputes over treaty interpretation, coverage or arbitrability go to the merits and are subject to deferential review under the New York Convention. Once jurisdiction is established, courts do not revisit arbitral determinations of scope or arbitrability de novo, but instead apply the Convention’s limited standards of review. Read together, Chevron, Stileks and Hulley reflect a consistent doctrinal framework: courts decide jurisdictional consent independently, but defer to arbitral tribunals on the merits once consent is established.

The Stabil District Court decision

The District Court’s decision in Stabil demonstrates how the Hulley framework operates at trial level. The petitioners, Ukrainian companies whose assets in Crimea were expropriated following Russia’s annexation, sought confirmation of a Swiss-seated arbitral award rendered under the Russia–Ukraine Bilateral Investment Treaty (BIT). Russia moved to dismiss, asserting sovereign immunity and arguing that no arbitration agreement had ever been formed with the petitioners.

Russia did not dispute that the BIT contains a standing offer to arbitrate disputes with Ukrainian investors. Instead, it argued that the offer never extended to the petitioners because they invested in Crimea when it was still part of Ukraine and Russia’s de facto control of Crimea does not convert the petitioners into covered investors under the BIT. According to Russia, the BIT could only function in mutually recognised territory, and therefore, no agreement to arbitrate was ever formed with respect to the petitioners’ claims because the international status of Crimea is disputed.

The District Court rejected this characterisation. The Court called Russia’s argument ‘rather cynical’ and ‘too clever by half’.^[7] Applying DC Circuit precedent, the Court held that Russia was not challenging the existence or validity of the arbitration agreement, but only its scope. The Court emphasised that, for FSIA jurisdictional purposes, the relevant inquiry is whether the tribunal purported to act pursuant to an arbitration agreement, not whether the tribunal ultimately applied the treaty correctly. Russia’s arguments concerning territorial reach and investor coverage, the Court concluded, conflated the FSIA jurisdictional inquiry with merits defences under the New York Convention.

Stabil on appeal: Russia’s jurisdictional theory under scrutiny

On appeal, Russia renewed its jurisdictional challenge by characterising the dispute as one of non-formation rather than scope. Russia argued, as it had in the District Court, that the Russia–Ukraine BIT never generated a valid offer to arbitrate with claimants such as Stabil because, in its view, the treaty could not extend to Crimea or to investments located there. On that theory, no arbitration agreement was ever formed, and the FSIA arbitration exception could not apply.

During the oral argument on 4 December 2025, the panel’s questioning focused on the consequences of that framing.^[8] Judges pressed Russia’s counsel on whether the argument effectively conceded the existence of a standing offer to arbitrate under the BIT while disputing only its application to a particular category of investors or territory. Several questions focused on whether Russia’s position would force courts to decide key treaty-interpretation issues at the jurisdictional stage, rather than later on the merits under deferential review.

The Court also tested the limiting principle of Russia’s theory. Judges asked whether, under Russia’s approach, virtually any disagreement about investor status, territorial scope or treaty coverage could be recast as a formation defect and used to defeat jurisdiction at the outset. That concern reflected a recurring theme in DC Circuit FSIA jurisprudence: while courts must decide de novo whether sovereign consent to arbitrate exists, they must also prevent the jurisdictional inquiry from swallowing the merits.

Viewed in that light, the oral argument reinforced the doctrinal boundary articulated in Chevron, Stileks and Hulley. Formation remains a judicial question. But once a treaty-based offer to arbitrate is established, disputes over how that offer applies belong to the merits, where judicial review is narrow and deferential. The Court’s questioning in Stabil suggests continued scepticism toward efforts to blur that line on appeal.

Practical implications for enforcement strategy

The emerging jurisprudence carries important implications for practitioners. Sovereign respondents can be expected to continue framing defences in terms of non-formation to trigger dismissal under the FSIA. Claimants, in turn, should be prepared to demonstrate the existence of a standing offer to arbitrate and the tribunal’s reliance on that offer, while resisting premature merits litigation at the jurisdictional stage.

At the same time, the Hulley decision introduces practical uncertainties. By requiring district courts to independently determine whether an arbitration agreement exists, Hulley may impose additional costs and delays on enforcement proceedings. As the DC Circuit acknowledged, this independent review cannot be outsourced to the arbitral tribunal, even where the tribunal conducted extensive proceedings on jurisdictional questions. Claimants should therefore anticipate that formation-based challenges will receive closer judicial scrutiny than scope-based arguments, and should be prepared to present evidence and expert testimony on foreign law questions that bear on whether a sovereign’s consent was validly given. Interestingly, the Hulley court suggested that the foreign court’s judgment enforcing the award may resolve the issue of the arbitration agreement’s validity based on the issue preclusion doctrine, though this approach in the FSIA context remains untested.

For courts, the challenge lies in maintaining the balance struck by the DC Circuit: safeguarding sovereign immunity where consent is genuinely absent, while preventing the FSIA from becoming a vehicle for wholesale re-litigation of arbitral jurisdiction and treaty interpretation.

Conclusion

Recent US decisions reflect a coherent and increasingly disciplined approach to arbitration-related jurisdiction under the FSIA. Courts insist on an independent judicial determination of whether an arbitration agreement was formed, but they are equally vigilant in distinguishing formation from scope. The Stabil District Court decision and the Court of Appeal’s oral argument confirm that this distinction is not merely theoretical but actively shapes outcomes at district court level. For the international arbitration community, the lesson is clear: consent remains the cornerstone of sovereign arbitration, but its determination turns on careful attention to contract-formation principles and the proper limits of jurisdictional inquiry.