Written by Dr. Nima Nasrollahi-Shahri (Sorbonne Law School) and Vincent Bassani-Winckler (PhD Candidate, Sorbonne Law School), both authors participated in the Working Group.

A few days ago, the Sorbonne Law School released the final report of a collective research project chaired by Professors Mathias Audit and Sylvain Bollée, entitled “Towards an EU Law on International Commercial Arbitration?”.

Conducted within the IRJS (Institut de Recherche Juridique de la Sorbonne), and more specifically its research group on private international law, SERPI (Sorbonne – Étude des Relations Privées Internationales), this project sets out to examine whether and how to improve the relationship between commercial arbitration and EU law.

Aims of the project and content of the report

Rather than proposing a full-scale harmonisation, the group focused on identifying limited and concrete modifications, focused on procedural issues, that would improve clarity, consistency, and the mutual recognition of arbitration-related judgments across Member States. Most notably, the report contains a proposal to qualify the arbitration exclusion in the Brussels I recast regulation and to add several provisions granting jurisdiction to the court of the seat of the arbitration, giving priority to these courts to prevent forum shopping and allowing arbitration-related judgments to circulate automatically within the EU.

The report is divided into three main parts. The first part of the report maps out the fragmented legal landscape currently governing international commercial arbitration within the European Union. Although arbitration is expressly excluded from the scope of the Brussels I Recast Regulation and Rome I regulation, it is not entirely isolated from EU law. For instance Regulation 2015/848 on insolvency proceedings refers to the effects of insolvency on pending arbitral proceedings, effects solely governed by the lex loci arbitri. By contrast, the jurisprudence of the CJEU has had a more substantial impact on arbitration-related matters, whether it is on application of EU public policy in arbitration (Mostaza Claro and Eco-Swiss) or of course investment arbitration between EU Member States (Achmea, Komstroy, and PL Holdings rulings). The CJEU has also shaped the scope of the arbitration exclusion in the Brussels I system. While early cases seemed fairly uncontroversial, West Tankers precluded Member States’ courts from issuing anti-suit injunctions relating to arbitration. Particularly controversial was the London Steamship Judgement, in which the Court limited the ability of a (then) Member State to refuse recognition of a judgment on the basis of a prior arbitration award – even where the award had already been confirmed by a court in that Member State (where the seat of arbitration was located).

The second part of the report lays out the rationale behind the working group’s proposals. It begins by acknowledging the political and legal constraints of a full-scale harmonisation, before arguing that targeted integration of arbitration-related rules into EU law – in particular the Brussels I Recast Regulation – would meaningfully enhance legal certainty, coherence, and the effectiveness of commercial arbitration within the Union. The report identifies a series of concrete legal issues where the current exclusion of arbitration from Brussels I Recast creates legal uncertainty or unfair outcomes. The first issue is certainly the risk of competing proceedings: the current framework does not give any priority, where the validity or applicability of an arbitration agreement is contested, to the judge of the seat of arbitration. Uncertainties remain, additionally, regarding the leeway of a judge of a Member State faced with a judgment rendered on the merits by the judge of another Member State after the latter has dismissed an arbitration agreement. Litigation concerning the constitution of the arbitral tribunal can also give rise to procedural conflicts. The circulation of decisions on the constitution of the arbitral tribunal and relating to the validity of the award are currently governed by a patchwork of national laws. Both could be ensured by a European recognition regime. In the wake of the London Steamship ruling the handling of conflicts between judgments and awards has never been more uncertain. In short, the current regime gives no clear priority to the court of the seat of arbitration, nor does it offer sufficient predictability to parties who rely on arbitration within the European judicial area.

In the final part of the report, the working group sets out a targeted reform plan for the Brussels I Recast Regulation. These proposed amendments are designed to strengthen the effectiveness of arbitration within the EU judicial area without harmonising the substance of arbitration law. Each provision responds to existing legal uncertainties or procedural inconsistencies and aims to enhance predictability, mutual trust, and party autonomy.

The proposed amendments to the Brussels I Recast Regulation

The amendments focus on six areas:

1.  Limited extension to arbitration of the scope of application of the Regulation (Article 1(2)(d))

Proposed provision (art. 1(2)(d)):

“This Regulation shall not apply to: (…) (d) arbitration, save as provided for in Articles 25 bis, 31 bis, 45 1. (d) and 45 3”

The first proposed amendment refines the current exclusion of arbitration from the Brussels I Recast Regulation. Presently, Article 1(2)(d) excludes arbitration entirely, which has led to interpretive tensions when arbitration-related issues intersect with judicial proceedings. The proposed reform retains the general exclusion but introduces narrowly defined exceptions – specifically for (proposed) Articles 25 bis, 31 bis, 45(1)(d), and 45(3).

This opening is not meant to harmonise arbitration law within the EU, but rather to create bridges where interaction with judicial mechanisms is unavoidable. It provides gateways for EU procedural law to engage with arbitration in discrete and functional ways, particularly around jurisdictional conflicts, enforcement of judgments, and safeguarding the role of the arbitral seat. Crucially, this shift does not introduce EU-wide arbitration rules. Instead, it merely extends the scope of the Regulation in a way that strengthens procedural consistency while continuing to respect the autonomy of Member States in substantive arbitration matters.

2. Recognition of Judgments Related to Arbitration (Article 2)

Proposed provision (art. 2):

“For the purposes of this Regulation: (a)(…) (…)

For the purposes of Chapter III, ‘judgment’ includes a judgment given by virtue of Article 25 bis paragraph 1 in the Member State where the seat of arbitration is located. It also includes a judgment given by virtue of Article 25 bis paragraph 1 (a) in another Member State, the court of which was expressly designated by the parties. It does not include a judgment issued by the court of another Member State on matters referred to in Article 25 bis paragraph 1; (…)”

This reform targets a critical gap in the existing system: the inability of arbitration-related court judgments (e.g. those concerning the annulment or enforcement of arbitral awards) to circulate within the EU under the automatic recognition regime of the Brussels I Recast.

The proposal amends Article 2 to include within the definition of “judgment” those decisions rendered either by the courts of the seat of arbitration (under Article 25 bis) or by courts expressly designated by the parties. Such judgments would now benefit from the mutual recognition mechanism of Chapter III. Conversely, judgments by other courts, not falling under these categories, would be excluded from automatic recognition.

This shift would enable decisions such as annulment or enforcement of awards issued by courts at the arbitral seat to circulate seamlessly across Member States. In effect, it creates a “European passport” for arbitration-related judicial decisions – enhancing legal certainty and mutual trust –  and preventing inconsistencies where one Member State’s court upholds an award and another ignores or contradicts it.

Importantly, this proposal, read in conjunction with article 25 bis, also ensures that parties retain freedom: they may still seek enforcement under national rules of jurisdiction if they prefer (art. 25, 3.). The reform merely introduces a uniform recognition track, based on mutual trust, building on the legitimacy of decisions from the arbitral seat.

3. Jurisdiction of the Courts of the Seat of Arbitration (Article 25 bis)

Proposed provision:

Article 25 bis:

“1. If the parties, regardless of their domicile, have agreed to settle their dispute by arbitration with its seat in the territory of a Member State, the courts of that Member State shall have jurisdiction over the following actions:

(a) Actions relating to the support for the constitution of the arbitral tribunal or the conduct of the arbitration procedure. This should be without prejudice to the jurisdiction of any other court expressly designated by the parties;

(b) Actions relating to the existence, validity or enforceability of the arbitration agreement. This should be without prejudice to:

• provisions of the national law of that State Member empowering the arbitral tribunal to rule on its own jurisdiction and, as the case may be, recognising it a priority in this respect; and
• article 31 bis paragraph2.

(c) Actions for annulment, recognition or enforcement of the arbitral award.

2. Actions referred to in paragraph 1 (a) and (b) may not be brought before a court of a Member State on the basis of national rules of jurisdiction.

3. Paragraph 1 (c) should be without prejudice to the right for a party to seek recognition and enforcement of an arbitral award before a court of a Member State on the basis of its national rules of jurisdiction.

4. The provisions of this article are without pre judice to the application of a rule of national law of the Member State where the seat of arbitration is located enabling the parties to waive their right to bring an action for annulment.

5. The provision of this article do not apply in disputes concerning matters referred to in Sections 3, 4 or 5 of Chapter II.”

This core reform introduces a new jurisdictional rule under EU law that recognises the centrality of the seat of arbitration. Under the proposed Article 25 bis, when parties have agreed to seat their arbitration in the territory of a Member State, the courts of that State will have jurisdiction over three key types of actions:

• (a) Requests for judicial assistance, such as the appointment of arbitrators;
• (b) Challenges to the existence, validity, or enforceability of the arbitration agreement; and
• (c) Actions for annulment, recognition, or enforcement of the award.

However, this is not a rule of exclusive jurisdiction in all cases. While Article 25 bis bars recourse to national jurisdiction rules for actions falling under (a) and (b), paragraph 3 expressly preserves the right for parties to seek enforcement of arbitral awards before other Member State courts, under those States’ existing national jurisdiction rules. In other words, a party could still apply directly for enforcement in a Member State other than the seat — which remains particularly important in practice for seeking execution against assets wherever they are located.

What this rule achieves, then, is not exclusivity per say, but a harmonised baseline: it grants primary jurisdiction to the courts of the seat for core functions, while preserving flexibility where appropriate. It also enhances coherence and foreseeability, notably by ensuring that judgments rendered by the court of the seat (especially on annulment or validity of awards) will benefit from automatic circulation under Chapter III of the Brussels I Recast (which is the effect of the proposed addition to article 2 (a)) — effectively granting them a “European passport.”

In addition, the rule accommodates Member States’ domestic doctrines, such as competence-competence and its negative effect, and waiver of annulment actions, making it fully compatible with diverse national legal cultures.

4. Priority of the Seat’s Courts in Conflicting Proceedings (Article 31 bis)

Proposed provision:

Article 31 Bis:

“1. Where a court of a Member State is seized of an action and its jurisdiction is contested on the basis of an arbitration agreement establishing the seat of the arbitration in another Member State, it shall, on the application of the party seeking to rely upon the said agreement, stay the proceedings until the courts of this other Member State have ruled or may no longer rule on the existence, validity or enforceability of the arbitration agreement.

2. However the court whose jurisdiction is contested continues the proceedings if:

(a) the arbitration agreement is manifestly inexistent, invalid or unenforceable under the law of the Member State where the seat is located; or

(b) the arbitral tribunal was seized and declined jurisdiction, and the arbitration agreement is inexistent, invalid or unenforceable under the law of the Member State where the seat is located.

For the purposes of this paragraph, reference to the law of the Member State where the seat is located encompasses conflict-of laws rules applicable in that Member State.

3. The provisions of this article are without prejudice of the application of a rule of national law of the Member State where the seat of arbitration is located empowering the arbitral tribunal to rule on its own juris diction and, as the case may be, recognizing it a priority in this respect.”

This reform introduces a stay mechanism to prevent jurisdictional races and forum shopping when disputes arise about the validity of an arbitration agreement.

When a court in one Member State is seized and the arbitration agreement designates a seat in another, the seized court must stay its proceedings until the courts of the seat have ruled — unless:

• The arbitration agreement is manifestly invalid, or
• The arbitral tribunal has already declined jurisdiction.

This reform addresses the recurring problem of inconsistent rulings and tactical litigation, where parties rush to court in jurisdictions likely to undermine arbitration. The proposed rule:

• Respects the primacy of the seat in deciding the validity of the arbitration agreement;
• Integrates negative effect competence-competence where national laws so provide (see para. 3);
• Ensures minimal interference by requiring only a prima facie validity to continue proceedings, thus filtering abusive challenges;
• Maintains consistency with the New York Convention, especially Article II(3), by offering a more favourable approach (per Article VII).

In practice, this rule harmonises procedural treatment of arbitration agreements across the EU and strengthens the parties’ contractual choices, giving effect to their selection of the arbitral seat as the appropriate forum for judicial review.

5. Clarification on Provisional Measures (Article 35)

Proposed provision:

Article 35: “Application may be made to the courts of a Member State for such provisional, including protective, measures as may be available under the law of that Member State, even if the courts of another Member State or an arbitral tribunal have jurisdiction as to the substance of the matter.”

This is a seemingly modest, but practically important clarification. Currently, Article 35 allows courts to grant provisional measures even if they lack jurisdiction on the merits — but it does not expressly mention arbitration.

The proposal amends this article to state that courts may issue such measures even if an arbitral tribunal has jurisdiction over the dispute. This codifies the approach taken by the ECJ in Van Uden.

6. Refusal of Recognition in Case of Conflict with Arbitral Awards (Article 45)

Proposed provision:

Article 45:

“1. On the application of any interested party, the recognition of a judgment shall be refused:

(…)

(d) if the judgment is irreconcilable with an earlier judgment given in another Member State or in a third State, or an arbitral award, involving the same cause of action and between the same parties, provided that the earlier judgment or arbitral award fulfils the conditions necessary for its recognition in the Member State addressed; or (…)

3. Without prejudice to point (e) of paragraph 1, the jurisdiction of the court of origin may not be reviewed. The test of public policy referred to in point (a) of paragraph 1 may not be applied to the rules relating to jurisdiction, including the rules governing the existence, validity or enforceability of arbitral agreements.”

This reform targets one of the most pressing weaknesses exposed by the London Steamship case: under current law, an arbitral award cannot itself prevent the recognition of a conflicting court judgment within the Brussels I framework.

The proposed change adds arbitral awards to the list of prior decisions that can bar recognition of later inconsistent judgments, provided that:

1. The award was rendered before the judgment,
2. Both involve the same cause of action and parties, and
3. The award meets the conditions for recognition in the requested state.

This ensures that awards enjoy the same res judicata value as earlier judgments, preventing inconsistent decisions and protecting the authority of arbitration.

In addition, paragraph 3 of Article 45 is revised merely to extend the prohibition of the use of public policy exceptions to the rules relating to jurisdiction, even when the rules governing the existence, validity or enforceability of arbitral agreements are at stake.

Conclusion: A Coherent and Functional Reform

These proposals are carefully calibrated. They do not seek to harmonise the substance of arbitration law in the EU – something neither realistic nor desirable given the diversity of legal traditions. Rather, the proposals aim to:

• Close procedural loopholes in the Brussels I Recast Regulation;
• Ensure legal certainty in cross-border litigation involving arbitration;
• Support party autonomy and reward the choice of a Member State seat;
• Enhance the attractiveness of European arbitration venues, through mutual trust in court supervision and support for arbitration.

In short, the proposals promote integration without harmonisation. They offer a modest but meaningful step towards a more coherent and predictable European framework for arbitration—one that recognises both the autonomy of arbitration and the importance of judicial cooperation in the EU.