Written By: Apoorv Agarwal, Manav Goyal
I. Introduction
A party that chooses arbitration is choosing to be done with the dispute faster. Unlike courts, arbitration is supposed to give a definitive answer quickly, without years of appeals. The arbitral award is meant to be the end of the road.
In October 2024, the Ministry of Law and Justice released the Draft Arbitration and Conciliation (Amendment) Bill, 2024 (the “Draft Bill”) for public consultation.[1] The Draft Bill proposes several changes to the Arbitration and Conciliation Act, 1996 (the “Act”). Many of these changes are welcome. But one proposal has attracted significant attention and debate i.e., the creation of an Appellate Arbitral Tribunal (“AAT”) under a proposed Section 34-A. This article examines what the AAT actually is, what problems it tries to solve, and whether it will succeed.
II. The Problem the AAT Is Trying to Solve
Under Section 34 of the Act, a losing party can approach a court and ask it to set aside an arbitral award. The grounds for doing so are limited. The award cannot be set aside merely because the court disagrees with the outcome. But in practice, Section 34 petitions have become a major source of delay. As of September 2023, over 2,100 Section 34 petitions were pending before the Delhi High Court alone, with each petition taking an average of over three and a half years to be decided.[2]
The law prescribes narrow grounds for challenge, but the process of litigating those grounds has become slow and expensive. This defeats the very purpose of arbitration. The Draft Bill tries to address this by offering parties an alternative: instead of going to a court under Section 34, they can choose to go to an AAT, which is a second arbitral tribunal constituted by an arbitral institution.
III. What Does the AAT Actually Do?
Before anything else, it is important to be clear about what the AAT is and what it is not. The word “appellate” suggests a full review of the award on merits. That is not what the AAT does.
The AAT can only examine the award on the same grounds that a court examines it under Section 34, broadly, procedural unfairness, excess of jurisdiction, or conflict with public policy. It cannot re-examine the facts or substitute its own view of the dispute. It is not a true appeal. It is, in substance, a replacement for the court under Section 34, not an additional tier on top of it.[3]
Under the current Act, a challenge to an award goes to a court under Section 34. If the Draft Bill is enacted, parties who have opted for an institutional arbitration will have a choice, they can go to the court under Section 34, or they can go to the AAT. The two routes are mutually exclusive. If parties opt for the AAT, the court’s jurisdiction under Section 34 is barred.[4]
After the AAT decides, either party can still file a statutory appeal under Section 37 before the High Court. So the total number of stages remains the same, the AAT simply replaces the court at one of those stages.
IV. Is the Concept New?
Two-tier arbitration where a first arbitral award can be challenged before a second arbitral tribunal is not a new concept in India. The Supreme Court upheld such a clause in M/s Centrotrade Minerals & Metal Inc. v. Hindustan Copper Ltd.[5] The Court held that there is nothing in the Act that prevents parties from agreeing to a procedure under which an arbitral award may be reconsidered by another arbitral tribunal. Party autonomy permits this.
What the Draft Bill does is different: it converts this contractual option into a statutory institution. It gives arbitral institutions the power to create AATs under their own rules, and it provides a statutory basis for such tribunals to function. The Draft Bill also limits the AAT to institutional arbitrations, ad hoc arbitrations are excluded.[6]
V. Problems with the AAT as Proposed
The idea behind the AAT is reasonable. The execution, however, raises serious concerns.
A. Institutional Bias: The Draft Bill proposes that the same institution which administered the original arbitration will constitute the AAT. For example, if an award is made under the Delhi International Arbitration Centre (DIAC), the AAT reviewing that award will also be constituted by DIAC. This creates a genuine appearance of bias. An institution reviewing the award of its own tribunal may, consciously or not, be inclined to protect it.[7] The Draft Bill does not address this concern.
B. No Clarity on Composition or Timelines: The Draft Bill does not say how many arbitrators the AAT should have, how they should be appointed, or within what time the AAT must decide. These are not minor details. Without clear rules, disputes over the constitution of the AAT will themselves require litigation, adding delay rather than reducing it.
C. The Section 36 Problem: Section 36 of the Act deals with enforcement of arbitral awards and the stay of enforcement during challenge proceedings. Section 36 currently vests jurisdiction exclusively in courts. The Draft Bill does not amend Section 36. So if a party goes to the AAT under Section 34-A, the question of stay of enforcement will still have to go to a court. This means parallel proceedings before two separate forums, a problem the Draft Bill was meant to solve.[8]
D. Only for Institutional Arbitrations: The AAT will only be available where parties have opted for institutional arbitration. A large number of arbitrations in India are conducted ad hoc i.e., without institutional administration. These arbitrations remain subject to the court-centric Section 34 process. This creates a two-track system where similarly situated parties are treated differently depending on whether their arbitration was institutional or ad hoc.
VI. What the Draft Bill Gets Right
The Draft Bill is not without merit.
First, the statutory recognition of emergency arbitration is a significant step. Until recently, Indian law had no provision for emergency arbitrators, interim relief before the constitution of the main tribunal had to be sought from courts. The Supreme Court’s decision in Amazon.com NV Investment Holdings LLC v. Future Retail Ltd.[9] recognised emergency arbitrators’ orders as enforceable, but only through judicial interpretation. The Draft Bill now proposes to give them express statutory recognition under a proposed Section 9-A. This removes ambiguity and makes India more aligned with international practice.
Second, the Draft Bill proposes to transfer certain powers from courts to arbitral institutions, specifically, the power to extend the time for making an award, to reduce arbitrators’ fees where delay is the tribunal’s fault, and to substitute arbitrators. These powers currently vest only in courts, which means even routine procedural issues require court intervention. Moving these powers to institutions is a practical and sensible reform.[10]
Third, the Draft Bill introduces a 60-day time limit for courts to decide applications under Section 8 (reference to arbitration). This is a targeted fix for a documented problem and does not require any structural change to the Act.[11]
VII. A Better Way to Fix Section 34
The AAT proposal is based on a correct diagnosis i.e., Section 34 petitions take too long. But the remedy is not necessarily to replace the court with a tribunal. A more direct approach would be to fix the Section 34 process itself.
Specifically, two things would help most. First, a mandatory time limit for courts to decide Section 34 petitions, similar to the 60-day limit proposed for Section 8. Second, clearer statutory language on what “patent illegality” and “public policy” mean in the context of Section 34, so that courts do not re-examine the facts of the dispute under the guise of judicial review.
The courts have themselves tried to discipline this. In Ssangyong Engineering & Construction Co. Ltd. v. National Highways Authority of India[12], the Supreme Court held that patent illegality must go to the root of the matter and does not permit re-appreciation of evidence. But lower courts have not always followed this consistently. Codifying this principle clearly in the statute would achieve more than creating a new tribunal.
If the legislature still wants to retain the AAT, the minimum changes needed are: mandatory appointment of AAT members from a panel maintained by the Arbitration Council of India (not by the institution that administered the original arbitration), a fixed time limit for AAT proceedings, an amendment to Section 36 so that stay of enforcement can also be sought from the AAT, and a defined limitation period for filing before the AAT. None of these appear in the current Draft Bill.
VIII. Conclusion
The Draft Arbitration and Conciliation (Amendment) Bill, 2024 is an ambitious attempt to reform India’s arbitration law. Many of its proposals, on emergency arbitration, on institutional powers, on time limits are genuinely useful. The AAT, however, is a complex structural change that has been proposed without working through the details. As drafted, it leaves too many questions unanswered on composition, bias, timelines, and its interaction with Section 36. The goal of arbitration reform should be to make the first decision final and not to create a faster route to a second decision. If the AAT is to be retained, it needs to be redesigned with greater care. Until then, the more effective path is to reform the Section 34 process within the courts rather than
[1]Ministry of Law and Justice, Government of India, Draft Arbitration and Conciliation (Amendment) Bill, 2024, released for public consultation on 18 October 2024.
[2]As of September 2023, 2,106 petitions under Section 34 were pending before the Delhi High Court, with average disposal taking approximately 1,327 days (over 3.5 years) and an average of 23 hearings per petition. See Amer Vaid, ‘Section 34 of Arbitration Act and Timely Disposal: Two Roads That Never Meet’, Bar & Bench (2023).
[3]Draft Bill, proposed Section 34-A(1). The AAT’s powers mirror the grounds available under Section 34 of the Act; it does not conduct a merits review of the award.
[4]Draft Bill, proposed amendment to Section 34(1) and the proviso thereto. The two routes, court under Section 34 and AAT under Section 34-A, are mutually exclusive.
[5]M/s Centrotrade Minerals & Metal Inc. v. Hindustan Copper Ltd., (2017) 2 SCC 228 (three-judge bench upholding validity of two-tier arbitration clause); final proceedings reported at (2020) 19 SCC 197. The 2006 Division Bench had initially held such clauses invalid: (2006) 11 SCC 245.
[6]Draft Bill, proposed Section 34-A(2). The AAT mechanism is restricted to institutional arbitrations only. Ad hoc arbitrations remain subject to the court-centric Section 34 mechanism.
[7]Draft Bill, proposed Section 34-A. The institution that administered the original arbitration is also empowered to constitute the AAT. The independence concern arising from this overlap is noted by several commentators. See Appellate Arbitral Tribunals under Arbitration (Amendment) Bill, 2024: Issues and Challenges, Mondaq (India), 27 August 2025.
[8]The Draft Bill does not amend Section 36 of the Act, which governs stay of enforcement of arbitral awards. A party that has opted for the AAT under Section 34-A will still need to approach a court for a stay order, creating parallel proceedings.
[9]Amazon.com NV Investment Holdings LLC v. Future Retail Ltd., (2022) 1 SCC 209. The Supreme Court held that an Emergency Arbitrator’s order, in an arbitration seated in India governed by institutional rules (SIAC), is enforceable as an order of the arbitral tribunal under Section 17(2) of the Act.
[10]Draft Bill, proposed amendment to Section 29-A. The powers to extend time for making an award, reduce arbitrators’ fees, and substitute arbitrators currently vested exclusively in courts are proposed to be delegated to arbitral institutions.
[11]Draft Bill, proposed amendment to Section 8. A 60-day time limit is proposed for courts to decide applications for reference to arbitration.
[12]Ssangyong Engineering & Construction Co. Ltd. v. National Highways Authority of India, (2019) 15 SCC 131. The Supreme Court held that patent illegality under Section 34(2A) must go to the root of the matter and does not permit re-appreciation of evidence.
