Modification, Merits, and Mayhem: Can Indian Arbitration Regain Coherence? (Part II)
In the first part of this piece, we traced the doctrinal foundations of Gayatri Balaswamy, examining how the judgment unsettles the balance between arbitral autonomy and judicial oversight by introducing the power of modification without clear statutory grounding. We argued that the Court’s attempt to distinguish modification from annulment collapses in practice, leaving unresolved the tension between merits review and the limited grounds under S.34.
This second part turns to the implications of that framework. It analyses how the Court’s reasoning has generated contradictions within its own formulation, blurred settled boundaries between modification and partial setting aside, and raised wider policy concerns about efficiency, finality, and party autonomy. It also explores unstructured alternatives under S.34(4), and the uneasy coexistence of modification with the Supreme Court’s extraordinary powers under Article 142. In doing so, it highlights how Gayatri Balaswamy leaves behind not doctrinal clarity, but a series of unresolved ambiguities as discussed below.
1. A lesser power to ‘partially set aside’ an award
In this dichotomy of modification and setting aside, there is a middle ground where the Court can ‘partly set aside’ a vitiated element if it is distinguishably separable from the remaining award. This is one of the key issues addressed in Gayatri Balaswamy. Though this power is recognized in decisions such as National Highways Authority of India v Trichy Thanjavur Expressway Limited (2023), it is yet to receive statutory recognition under the Arbitration Act.
The grounds to exercise this power is recent but courts have applied it clearly. The power to partially separate and set aside an offending part of the award would depend on whether that part is ‘independent and distinct’ and whether the removal of that specific part would disrupt or impact any other part of the award decided by the Tribunal. The question of severability would have to be decided bearing in mind whether the claims are interdependent or so intertwined that one cannot be segregated from the other. If the separation meets this standard, then that vitiated element of the award can be set aside. Recognizing this test, the 2024 T.S. Vishwanathan Report suggested amending S.34 to replace ‘set aside by the Court’ in sub-sections 2 and 2A with ‘set aside in whole or in part by the Court’ to allow partially setting aside an award by using severability.
While the authors agree with the existence and exercise of the power to partially setting aside as it is merely extension of the hands-off approach, however there is a convolution in the Court’s assessment that must be addressed.
In paragraph 85, the Court concludes its analysis by recognizing a limited power of modification under S.34 and S.37. It specifies four circumstances in which this power may be exercised. First, where an award is severable, the “invalid” portion may be separated from the “valid” portion. Second, clerical, computational, or typographical errors that are apparent on the face of the record may be corrected. Third, post-award interest may be modified in certain circumstances. Fourth, under Article 142 of the Constitution, the Court may exercise its inherent powers with care and caution. Thus, the Court seeks to carve out narrow situations where modification, as distinct from annulment, is permitted.
Within these four circumstances, the first circumstance in paragraph 85 exposes a critical fault line in the Court’s framework. On the one hand, the Court expressly acknowledges that modification and setting aside are conceptually distinct and mutually exclusive powers, while modification alters an award, setting aside annuls it altogether. (¶38) Yet in ¶85, the Court paradoxically subsumes severability within the power of modification to “sever the invalid portion from the valid portion.” This formulation blurs the very distinction the Court had earlier emphasised (at ¶38). Severability is already a recognised ground within S.34, distinct from modification, enabling courts to set aside the flawed portion without disturbing the rest of the award. To recharacterize this as a circumstance to permit modification collapses two distinct powers into one, generating confusion about whether modification constitutes annulment or alteration. The Court thus contradicts its own analysis, creating doctrinal uncertainty about the precise contours of modification vis-à-vis setting aside by both reading modification as a power which subsumes the power to partially setting aside (at ¶85) and as mutually exclusive powers at (¶38).
2. Policy Considerations in allowing Modification and Variance
Critical policy concerns arise when the scope of modification within the framework of setting aside arbitral awards is widened.
Firstly, the legislative scheme of S.34, explicitly limits its intent confining judicial interference to only jurisdictional or procedural defects. By introducing powers of modification, the intended divide between setting aside and appellate review is effectively negated. Instead of only assessing defects, courts can now address errors and correct deviations from the ideal interpretation. While this may act as a corrective measure, it conflicts with party autonomy. It is also time-consuming and delays the finality that arbitral awards are meant to provide.
Secondly, the Arbitration Act of 1940 expressly empowered courts to modify arbitral awards under S.15 (in challenge proceedings) and S.39 (in appeals). However, this power was subsequently omitted in the Arbitration and Conciliation Act, 1996. Both the Draft Amendment, 2024 and the T.S. Vishwanathan Report propose that courts may set aside an award “in whole or in part.” Yet, rather than allowing modification, they alternatively recommend remitting the award back to the arbitral tribunal for correction. The omissions and the recommendations collectively indicate that courts are not to modify arbitral awards but only to remit or set them aside under S.34.
Thirdly, since awards can still be appealed even after being modified, the possibility of further alteration encourages challenges. This is neither procedurally sound nor commercially efficient. ‘It is the nature of an arbitral award to get challenged.’ (Saurabh Kirpal, Gayatri Balaswamy, Submissions) Therefore, prolonged litigation may arise wherein one party seeks a subsequent modification (as seen in S.39 of the Arbitration Act, 1940), while the other resists it under S.5. Such modification proceedings can instigate legal uncertainty and commercial treachery, ultimately defeating the objective of arbitration to provide an expeditious and final award.
Therefore, a wide interpretation of S.34 wherein the power to modify is included, transcends the normative contours envisaged by the legislature while creating overarching policy implications.
3. An unstructured alternative under S.34 (4)
An alternative to modification is for courts to use their power under S.34(4). This provision gives the Court the power to adjourn the proceedings for a specific time period in order to enable the arbitral tribunal the opportunity to take steps to eliminate the grounds for setting aside the award. Superficially, this provision might appear to be a median which allows the tribunal i.e., the original body which delved into the merits, to cure the alleged defects in the award, whilst restricting the interference of the Court into the merits of the matter. But it reveals a fundamental doctrinal difficulty with respect to its harmonious interpretation with S.32 of the Act.
S.32 of the Act deems the proceedings to be terminated upon the passage of the final arbitral award. Upon the termination of proceedings, the arbitral tribunal becomes functus officio. Its mandate expires and along with it, its authority to revisit the substance of the dispute under S.32(3). This creates a clear paradox where the tribunal’s mandate has ended but it is still asked to exercise powers it no longer holds.
While an argument of ‘temporary revival’ of the tribunal can be made, the lack of structural and procedural safeguards to guide the temporal limits of the power under S.34(4) upon the tribunal raises concerns. One of the significant concerns relates to the lack of clarity with regards to the appropriate stage under S.34 proceeding of making this remittal, whether during the course of arguments, prior to the final determination of the dispute or final hearing. This indeterminacy has accelerated interpretative ambiguity and has in practice, incentivized parties and at times, the courts to stretch the provision to act a surrogate avenue for altering the substance of the award. This line of interpretation strikes at the edifice of arbitration i.e., finality and timely efficiency. A more structured approach mandating that remittal be considered at the earliest possible stage, preferably at the stage of admission or initial arguments, would enhance procedural efficiency and reduce avoidable delays.
4. 142: Constitutionally Settled but Commercially Dissonant – modification cannot prevent curative petitions and the use of extraordinary powers.
The Court’s conflation of modification under procedural review with the exercise of extraordinary powers reflect a doctrinal inconsistency between availing remedies u/s.34 or u/Art.142. The Supreme Court’s use of its extraordinary powers to overturn an award in a dispute has been the subject of much debate and criticism in recent legal scholarship as ‘excessively interfering’ and ‘legally erroneous’. Whether amending awards or curative petitions, the use of Art.142 is often critiqued to be antithetical to the Act, violative of party autonomy and finality. Herein, a curative petition lays down a multi layered appellate mechanism where post arbitration recourses first include a challenge u/s.34; followed by an appeal u/s.37; then a Special Leave Petition u/Art.136 which is then resolved u/Art.137, permitting the Supreme Court to review its own decisions and finally followed by a curative petition u/Art.142.
The five-judge constitutional bench in Gayatri Balaswamy had a critical opportunity to overrule or, at least, read down Delhi Metro Rail Corporation v Delhi Airport Metro Express Private Limited (2024) (“DMRC v DAMPEL”).
Despite possessing ample bench strength, the Gayatri Balaswamy Bench’s addressal of this issue remained superficial (¶82-84). The bench could have imposed objective tests, limitations or criterion with respect to limiting the SC extraordinary powers u/Art.142 or appealability of awards, even at the risk of judicial overreach as the courts have had an established practice to lay down such objective criterion in arbitration (here, here and here). Instead of establishing concrete standards, the court retreated into platitudes by merely citing Shilpa Sailesh v Varun Sreenivasan (2023) and stating that ‘this Court must be conscious’ (¶84) while exercising its power u/Art.142. Though in the wrong sense, the court reasoned that the exercise of Art.142 powers to modify awards would ‘end protracted litigation’ and ‘save parties’ money and time’ (¶84).
Regardless, this consideration arises as an afterthought rather than a foundational principle for a doctrinal framework. While the constitutional dust has settled in the exercise of Art.142, the commercial irregularity continues to exist. These concerns about time and cost must be addressed early under S.34 or S.37, not left to be resolved later under Art.142.
In this regard, it is firstly argued that there exists no direct nexus between the modification of awards and the appealability under constitutional remedies. Modification, even if limited to a procedural review, does not affect the right of an aggrieved party to invoke curative jurisdiction under Art.142. The assumption that modification inherently reduces protracted litigation is speculative as there is no bar preventing a dissatisfied party from appealing the award or filing a curative petition after the award’s modification. Therefore, the Court’s reasoning conflicts with two distinct remedies, failing to acknowledge that modification cannot operate as a legal substitute for constitutional rights and remedies.
Secondly, there is another inherent disconnection between the Court’s substitution of modification as a form of procedural correction and the exercise of Art.142 powers in arbitral matters. The Court itself acknowledges that modification is confined to procedural irregularities (¶50), yet precedents such as DMRC v DAMPEL and Central Organisation for Railway Electrification v ECI-SPIC SMO MCML (JV) (2025), demonstrate that invocation of Art.142 must extend to a full merits review to achieve ‘complete justice’ whether in curative petitions or otherwise. This contradiction exposes the logical fallacy in suggesting that a procedural remedy (modification) can pre-empt and substitute the need for a substantive constitutional remedy (Art.142). If both operate in distinct, mutually exclusive domains, as distinguished by the Court, one cannot logically be deployed to foreclose or reduce the other.
Concluding Remarks
In recent times, the Indian arbitral regime has been subject to rapid and conflicting developments. The Courts’ proclivity for expansive judicial interpretation into arbitration with conflicting judgements, as aforementioned, has introduced unnecessary complexity. This automatically leads to a cumbersome arbitral regime.
Indian institutional reluctance is becoming increasingly visible. The Public Works Departments in Delhi has deleted arbitration clauses from their standard contracts, signalling the state’s waning confidence from the arbitral process. On one hand the Chief Justice of India and the judiciary propagate arbitration to attract foreign investors and the private sector, whereas on the other the state simultaneously retreats from arbitral mechanisms in its own public procurement processes, reflecting a hypocritical stance overall. While India aspires to project itself as an arbitration-friendly jurisdiction, persistent challenges concerning arbitral finality, commercial expediency, and autonomy continue to undermine these efforts. Without correcting these foundational pillars, mere promotional efforts remain hollow.
The reluctance of the Gayatri Balaswamy Constitutional Bench to engage in correcting substantive issues is unsatisfactory. For India to emerge as a ‘hub of commercial arbitration’ the courts must limit their intervention and maintain arbitral sanctity to pave the way for the future of Indian arbitral regime.
About the Authors
Dhruv Madan and Sanjana Pershad are incoming penultimate-year law student at Jindal Global Law School. Animesh Bordoloi (LL.M., International Arbitration & Dispute Resolution, National University Singapore) is a Senior Counsel at Asian International Arbitration Centre.