FROM JUDICIAL IMPROVISATION TO STATUTORY CODIFICATION: A CRITICAL AND COMPARATIVE ANALYSIS OF EMERGENCY ARBITRATION IN INDIA, SINGAPORE, AND HONGKONG

I. INTRODUCTION AND RESEARCH PROBLEM

1. No statutory recognition — Despite two rounds of amendments to the Arbitration and Conciliation Act, 1996 (in 2015 and 2019), Indian arbitration law contains no statutory definition or recognition of “Emergency Arbitrator” or “Emergency Arbitration.”
2. The resulting fragmented choice — In the absence of statutory backing, parties seeking urgent interim relief before a tribunal is constituted face two options:
   • Approach Indian courts under Section9, or
   • Invoke institutional emergency arbitrator rules (SIAC, ICC, LCIA)
3. The enforcement gap — Orders passed by institutional emergency arbitrators have no express domestic enforcement mechanism under Indian law.
4. The judicial fix — Amazon (2021) — In Amazon.com NV Investment Holdings LLC v. Future Retail Ltd., the Supreme Court read an EA order as an order of the tribunal under Section17(1), making it enforceable under Section17(2). This resolved enforceability, but only for India-seated arbitrations.
5. The unresolved gap — Part I applies only to India-seated arbitrations (BALCO, 2012). The 2015 proviso to section2(2) extends sections 9, 27 and 37(1)(a) and (3) to foreign-seated international commercial arbitrations unless the parties agree otherwise, but it does not extend section17.
6. The core problem — The result is a doctrinally unstable, seat-contingent, judge-made regime standing in for what other leading arbitration jurisdictions have addressed through express statutory provisions.

II. THE TWO TRACK STRUCTURE OF INTERIM RELIEF

The Act provides two distinct tracks for interim relief. Section9 allows a party to approach a court, either before the commencement of arbitral proceedings, during the proceedings, or at any time before enforcement of the award. Section17 empowers the arbitral tribunal itself, once constituted, to grant equivalent interim measures, with such orders enforceable “in the same manner as if it were an order of the Court” following the 2015 Amendment. The difficulty lies precisely in the period between the commencement of arbitration and the constitution of the tribunal — a window during which Section17 relief is unavailable, and Section9 relief from a court, while available, may be slower, less confidential, and less attuned to the commercial context of the dispute than relief from an arbitrator familiar with the underlying transaction.

The 246th Report of the Law Commission of India, submitted in 2014, recommended that the definition of “arbitral tribunal” under Section2(1)(d) of the Act be expanded to include an emergency arbitrator appointed under institutional rules. Parliament did not act on this recommendation either in the 2015 Amendment or the subsequent 2019 Amendment. This legislative inaction is significant: it meant that by the time Amazon reached the Supreme Court, the Court was aware that Parliament had been alerted to the gap and had chosen not to close it. The opposing party in Amazon relied precisely on this legislative silence to argue that the Court should not read in a concept Parliament had declined to codify. The Supreme Court rejected this argument, reasoning that the absence of an express prohibition did not preclude a contextual reading of the statute that gave effect to party autonomy and the institutional rules the parties had themselves chosen.

III. THE INDIAN JUDICIAL TRAJECTORY

Before the Supreme Court’s intervention, Indian courts had already begun improvising indirect solutions, though none amounted to direct recognition of emergency arbitration. In HSBC PI Holdings (Mauritius) Ltd. v. Avitel Post Studioz Ltd., a SIAC-seated emergency arbitrator had granted interim relief freezing the respondents’ assets; the Bombay High Court did not enforce that emergency award as such, but independently granted a substantially similar order under Section9, effectively giving the emergency arbitrator’s reasoning persuasive, if not binding, force. Similarly, in Raffles Design International India Pvt. Ltd. v. Educomp Professional Education Ltd., 2016 SCC OnLine Del 5521, the Delhi High Court held that even where a foreign-seated emergency award could not be directly enforced under the Act, the successful party could independently approach an Indian court under section9 on the same facts. In Ashwani Minda v. U-Shin Ltd., 2020 SCC OnLine Del 1648, however, the applicant, refused relief by an emergency arbitrator under the JCAA Rules, sought the same relief under section9. The Delhi High Court held the petition not maintainable: the parties had excluded Part I, and a party that has elected the emergency arbitrator cannot use section9 as an appeal. These decisions show courts reaching for pragmatic workarounds, but workarounds that forced parties to re-litigate urgency afresh, defeating much of the speed and finality that emergency arbitration promises.

Amazon v. Future Retail

In 2019, Amazon.com NV Investment Holdings LLC invested in Future Coupons Pvt. Ltd., a group entity closely tied to Future Retail Ltd. (“FRL”), India’s second-largest offline retailer at the time. The shareholders’ agreement named the Reliance group among “restricted persons” to whom FRL’s retail assets could not be transferred without Amazon’s consent. In 2020, FRL nonetheless proceeded to sell its retail, wholesale, and logistics businesses to Reliance in a transaction valued at approximately ₹24,700 crore. Amazon treated this as a breach and invoked emergency arbitration under the SIAC Rules, as the arbitration agreement was seated in New Delhi but administered under SIAC’s institutional framework. The SIAC-appointed emergency arbitrator issued an award in October 2020 restraining FRL from proceeding with the transaction. FRL refused to treat the emergency award as binding, arguing that the Act does not recognise an “emergency arbitrator” within its definition of “arbitral tribunal,” and pressed ahead with regulatory approvals for the Reliance deal.

Amazon sought enforcement of the emergency award under Section17(2) before the Delhi High Court. A Single Judge upheld enforceability; a Division Bench stayed that order; the matter reached the Supreme Court on appeal. On 6 August 2021, the Supreme Court held that an emergency arbitrator’s award, in an India-seated arbitration conducted under institutional rules that expressly provide for such a mechanism, constitutes an order of the arbitral tribunal under Section17(1) and is enforceable as such under Section17(2). The Court grounded this conclusion in party autonomy: because the parties had voluntarily agreed to SIAC’s institutional rules, which expressly provide for emergency arbitration, they could not later disown the consequences of that choice merely because the Act itself was silent on the concept. The Court also drew on its earlier reasoning in Avitel, treating that decision’s willingness to grant substantively identical relief under Section9 as evidence that Indian courts had already implicitly accepted the legitimacy of emergency-arbitrator-style relief.

The judgment carries three distinct doctrinal consequences, which commentary on Amazon often conflates. First, an India-seated emergency order is directly enforceable without a fresh section9 application, a considerable efficiency gain over the Avitel and Raffles Design approach. Second, the Court held that an order enforcing an emergency arbitrator’s order under section17(2) is not appealable under section37 (discussed under Critique below). Third, and most consequential for this paper’s argument, the Court confined its holding to India-seated arbitrations. It left untouched emergency orders rendered in arbitrations seated outside India, which still require a fresh section9 application, the very workaround that Amazon’s reasoning was meant to make unnecessary.

IV. THE FOREIGN-SEATED GAP

The most significant unresolved problem is the treatment of foreign seated emergency awards. India is a signatory to the New York Convention, but the Convention itself, along with the UNCITRAL Model Law, does not define an emergency arbitrator’s order as an “award” for the purposes of cross-border recognition, leaving this a matter for domestic legislation to resolve. Because Amazon confined its ruling to India-seated arbitrations, a party that obtains a Singapore- or London-seated emergency award still cannot enforce it directly in India; it must instead approach an Indian court afresh under Section9, essentially retracing the pre-Amazon workaround for a large category of commercially significant cases, since many Indian parties in cross-border contracts agree to foreign seats. This is a doctrinal inconsistency that judicial interpretation, confined as it necessarily is to the facts before the court, cannot resolve on its own; only legislative amendment can extend enforceability to foreign-seated emergency awards, and even then, only in a manner consistent with India’s obligations and reservations under the New York Convention.

Critique

A further critique is distributional. Emergency arbitration under institutional rules such as SIAC’s is not a mechanism a small or medium enterprise can readily access: institutional fees for emergency arbitrator applications, together with the practical need for experienced international counsel, place it largely within reach of well-resourced corporate parties of the kind involved in Amazon itself. Smaller domestic parties, even where their arbitration agreement nominally permits emergency arbitration, remain practically dependent on the slower Section9 court route. This critique does not undermine the desirability of emergency arbitration as a mechanism, but it does suggest that codification exercises should be paired with efforts — for instance, through the Arbitration Council of India providing simplified emergency arbitrator panels for domestic institutional arbitration — to prevent the mechanism from developing into a two-tier system of urgent relief.

Because Amazon treats an emergency order as a section17 order, one would expect section37 to supply the appellate route. Yet the Court held that an order enforcing it is not appealable under section37. Section37 was drafted with a constituted tribunal’s Section17 orders in mind, not with the compressed, ex parte-heavy, and rules-driven procedure typical of emergency arbitration. Questions such as whether an appellate court reviewing an emergency award should apply the same standard of scrutiny as it would to a fully-heard Section17 order, or whether it should show greater deference given the emergency arbitrator’s necessarily abbreviated process, remain unaddressed by Amazon and by the Act as it currently stands. This is precisely the kind of procedural detail that is difficult for a court to supply through interpretation and is more naturally the province of a specifically drafted statutory provision.

V. COMPARATIVE PERSPECTIVE: SINGAPORE AND HONG-KONG

 Singapore and Hong Kong have both given emergency arbitration express statutory footing, in different ways. Singapore’s International Arbitration Act 1994, as amended in 2012, includes the emergency arbitrator within “arbitral tribunal” (s 2(1)), so emergency orders in Singapore-seated arbitrations are enforced like tribunal orders, and the High Court has treated foreign-seated emergency awards as enforceable foreign awards (CVG v. CVH). Hong Kong did not change its definition of the tribunal. It inserted Part 3A into the Arbitration Ordinance (Cap. 609) in 2013, and section22B makes emergency relief enforceable with leave of the court, whether granted in or outside Hong Kong. For foreign relief, leave is available only for listed temporary measures, such as preserving the status quo, assets or evidence, or giving security, and the leave decision is not appealable. India has no statutory recognition, and foreign-seated orders need a fresh section9 application.

First, both jurisdictions legislated before a landmark dispute forced the issue, whereas India’s regime emerged from a single high-value case. Second, Singapore adjusts definitions, which is simple but leaves cross-border questions to case law, while Hong Kong keeps the emergency arbitrator distinct and legislates only for enforcement. Third, Hong Kong shows that seat-neutrality is a policy choice, bounded by a closed list of measures. For India, a definitional change alone will not do, because the Act separates section9, section17 and Part II enforcement.

VI. THE DRAFT ARBITRATION AND CONCILIATION (AMENDMENT) BILL, 2024

Following a report submitted in February 2024 by an Expert Committee chaired by Dr T.K. Viswanathan, the Department of Legal Affairs released the Draft Arbitration and Conciliation (Amendment) Bill, 2024 for public consultation in October 2024. Among its central proposals is a new section 9A, which would introduce a formal statutory definition of “emergency arbitrator” and permit an arbitral institution to appoint one, before the constitution of the arbitral tribunal, for the purpose of granting interim measures of the kind ordinarily available under section 9. Orders passed by an emergency arbitrator under the proposed framework would be enforceable in the same manner as an order of the arbitral tribunal, and would be liable to confirmation, modification, or vacation by the tribunal once constituted.

This proposal has several evident merits. It converts a judicially-implied position into an express statutory entitlement, removing the interpretive uncertainty that necessitated litigation as far as the Supreme Court in Amazon. It also signals, in line with the Bill’s broader stated aim of strengthening institutional arbitration, that Parliament intends emergency arbitration to operate through recognised arbitral institutions rather than on an ad hoc basis, which should improve procedural consistency. However, the draft does not fully resolve the foreign-seated gap: the revised proviso to section 2(2) extends only section 9A(2), which concerns the conduct of emergency proceedings, to foreign-seated arbitrations, and not the enforcement provision in section 9A(3)

• Amazon: Enforceability Resolved, Doctrinal Fragility Entrenched

 By characterizing an emergency arbitrator’s order as an order “of the arbitral tribunal” under Section17(1), the Supreme Court did not merely interpret the Act — it stretched the definition of “arbitral tribunal” beyond what Section2(1)(d) contemplates, since an emergency arbitrator is functus officio the moment the actual tribunal is constituted and never adjudicates the substantive dispute.

Section2(1)(d) defines “arbitral tribunal” as a sole arbitrator or panel constituted to resolve the dispute referred. An emergency arbitrator’s mandate is definitional temporary and interim — a structurally different function. The Court’s equation of the two is a purposive reading justified by outcome (enforceability), not by textual fit. This matters because a purposive fix that strains statutory text is inherently less stable than an amendment — a future bench with a more textualist inclination could revisit it. The Court’s reasoning rested on party autonomy under sections 2(1)(a), 2(6), 2(8) and 19(2), not on the text of section2(1)(d). The Draft Bill confirms the fragility of the equation: it inserts a separate definition in section2(1)(ea) and enforces emergency orders “as if” they were section17(2) orders (proposed section9A(3)). Parliament’s own drafting treats the equation as a fiction to be written down, not a settled reading of the existing text.

• The foreign-seated gap reflects statutory design, not omission

Commentators often describe the foreign-seated gap as something Amazon “failed” to address. A sharper framing is that the gap is the product of a deliberate statutory choice, so only legislation, not judicial interpretation, can close it.

Bharat Aluminium Co. v. Kaiser Aluminium Technical Services Inc., (2012) 9 SCC 552, abandoned the earlier Bhatia International approach and held that Part I applies only to arbitrations seated in India (prospectively, from 6 September 2012). In 2015 Parliament qualified this through the proviso to section2(2): for international commercial arbitrations seated outside India, sections 9, 27 and 37(1)(a) and (3) apply unless the parties agree otherwise. Section17 is not on that list. Parliament therefore chose to give foreign-seated arbitrations access to Indian courts for interim relief, but not the section17(2) enforcement mechanism. A court extending section17 to foreign-seated emergency orders would have to read into the proviso a provision it omits, and would also have to depart from BALCO’s territorial principle. Amazon, which concerned an India-seated arbitration, had no occasion to do so. The gap can therefore be closed only by amending the proviso or adding a new provision, which is what the Draft Bill attempts, only in part.

• India’s silence on EA is inconsistent with its own stated New York Convention posture

India’s practical treatment of foreign EA orders (routing parties back to Section9, rather than treating foreign EA orders as arbitral awards or orders capable of recognition) sits awkwardly against India’s general policy commitment to a pro-enforcement, Convention-friendly arbitration regime.Whether an interim emergency order can be a “foreign award” under section44 is contested. The stronger view is that it cannot, because it does not finally determine any claim. A foreign-seated emergency order therefore sits between Part I and Part II

• Codification Risk

Over-legislating could reintroduce the very court-interference problem the 2015/2019 amendments tried to eliminate

A serious original contribution is acknowledging that badly drafted EA legislation could reopen the door to expansive judicial review of EA orders — recreating exactly the pre-2015 problem of excessive court interference that the amendments were designed to curb.

The 2015 amendment’s core purpose (per the 246th Law Commission Report) was minimizing court intervention. Any statutory EA provision would need a narrow, clearly bounded review standard (akin to Section37’s limited appeal grounds) — otherwise, giving courts a fresh statutory hook to scrutinize EA orders substantively could undo the “minimal interference” gains of the last decade.

VII. CONCLUSION

Amazon v. Future Retail was a necessary and, on balance, correctly reasoned judgment: faced with an evident gap between institutional practice and statutory text, the Supreme Court chose an interpretation that honoured party autonomy and prevented a party from invoking institutional rules only to disown their consequences once those rules proved inconvenient. But the judgment, precisely because it was compelled to work within the four corners of an existing statute never designed with emergency arbitration in mind, could not do everything that a purpose-built statutory provision could achieve. It resolved the enforceability question for India-seated emergency awards while leaving the foreign-seated question, the accessibility question, and the appellate-review question substantially open.

The Draft Arbitration and Conciliation (Amendment) Bill, 2024’s proposed Section9-A represents the right institutional response to this gap, converting the Amazon position into codified law and aligning India more closely with the approach long taken by Singapore. For the reform to deliver its full promise, however, three refinements are worth considering as the Bill moves toward its final form. First, the legislature should consider following Hong Kong’s example by extending statutory recognition to foreign-seated emergency awards, subject to appropriate safeguards, rather than confining the new Section9-A to institutions operating under Indian auspices. Second, the Arbitration Council of India should be tasked with developing accessible, lower-cost emergency arbitrator panels for domestic institutional arbitration, so that the benefits of the mechanism are not confined to large corporate disputants of the kind involved in Amazon itself. Third, the interaction between a Section9-A emergency order, its confirmation or modification by the constituted tribunal, and any subsequent appeal — whether under Section37 or the proposed Appellate Arbitral Tribunal — should be spelt out with greater procedural precision than the current draft provides, so that emergency relief in India does not merely trade judicial uncertainty for statutory ambiguity. Legislative codification, done well, can finally give India’s emergency arbitration regime the doctrinal stability that a single Supreme Court judgment, however sound, was never in a position to provide on its own.

VIII. REFERENCES

1. Arbitration and Conciliation Act 1996 (India), as amended in 2015, 2019 and 2021.
2. International Arbitration Act 1994 (Singapore) (2020 Rev Ed), ss 2(1), 12, as amended in 2012.
3. Arbitration Ordinance (Cap. 609) (Hong Kong), Part 3A, ss 22A, 22B.
4. Law Commission of India, 246th Report on Amendments to the Arbitration and Conciliation Act, 1996 (2014).
5. Expert Committee on Arbitration Law (T.K. Viswanathan Committee), Report (2024).
6. Department of Legal Affairs, Government of India, Draft Arbitration and Conciliation (Amendment) Bill, 2024.
7. Bharat Aluminium Co. v. Kaiser Aluminium Technical Services Inc., (2012) 9 SCC 552 (SC).
8. HSBC PI Holdings (Mauritius) Ltd. v. Avitel Post Studioz Ltd., 2014 SCC OnLine Bom 102 (Bom HC).
9. Raffles Design International India Pvt. Ltd. v. Educomp Professional Education Ltd., 2016 SCC OnLine Del 5521 (Del HC).
10. Ashwani Minda v. U-Shin Ltd., 2020 SCC OnLine Del 1648 (Del HC).
11. Amazon.com NV Investment Holdings LLC v. Future Retail Ltd., (2022) 1 SCC 209 (SC).
12. CVG v. CVH [citation to be completed] (Singapore High Court).
13. SIAC, Arbitration Rules (2010 and later editions), emergency arbitrator provisions.