Pragmatism over Pedantry: In Defence of the Power to Modify Arbitral Awards Post-Gayatri Balaswamy

Prabhas Kumar^[1] & Surya Prakash Swain^[2]

Introduction

The Supreme Court’s (“SC”) majority ruling in Gayatri Balaswamy v. ISG Novasoft Technologies Ltd. (2025) (“Gayatri Balaswamy”) holds that Indian courts may, in narrow circumstances, modify arbitral awards under Section 34 of the Arbitration and Conciliation Act, 1996 (“the Act” or “the 1996 Act”). The erstwhile Chief Justice Sanjeev Khanna, while writing for the 4:1 majority bench, reasoned that the existing recourses available to the court when an award is challenged essentially leave the parties at three difficult positions — (i) to have the award annulled in its entirety, (ii) severed and enforced in part, or (iii) be sent back for re-arbitration upon their consensus. This, as per the Majority, negates the raison d’être of the arbitral process, i.e., cost-effective and expeditious dispute resolution, and hence necessitates the Courts to “hammer out the creases” through a limited power of modification.

The dissenting opinion by Justice Viswanathan, now the cornerstone of most criticism against this landmark judgment, contends that the absence of an explicit statutory power to modify awards under the 1996 Act constitutes a clear bar on such judicial intervention and flags multiple concerns regarding this new power. The present article engages with and responds to the principal criticisms of the judgment, addressing each in turn, effectively demonstrating how the Apex Court has chosen pragmatism over pedantry while strengthening India's Arbitration Regime.

A Jurisprudence of Contradiction: The Road to Gayatri Balaswamy

To understand the monumental significance of the ruling in Gayatri Balaswamy, one must first journey through the jurisprudential wilderness that plagued the Indian arbitration regime for the last three decades.

The Arbitration Act of 1940 dominated the landscape preceding the present statute. It treated arbitral tribunals as junior partners in the judicial process. Its provisions, particularly Sections 15 and 16, armed courts with sweeping powers to "modify or correct" awards, effectively turning arbitration into a mere dress rehearsal for protracted court battles, and fostered a culture of excessive judicial meddling.

The 1996 Act, inspired by the principles of the UNCITRAL Model Law of 1985, was intended to be a revolutionary departure from this inconsistency. Its stated objective, "to minimise the supervisory role of courts in the arbitral process," was a clear legislative mandate to break from the interventionist past. The conscious excision of the word "modify" from its text was the boldest symbol of this new, pro-arbitration philosophy.

However, the ghost of modification, fuelled by judicial discomfort with absolute powerlessness in the face of injustice, refused to be banished. This led to a deeply schizophrenic jurisprudence, where the judiciary's actions often stood in stark contradiction to its stated principles. In McDermott International Inc v Burn Standard Co Ltd, the Court declared that it "cannot correct errors of the arbitrators," only to paradoxically modify the award on the question of interest by invoking its plenary powers under Article 142. Conversely, in Project Director, NHAI v M Hakeem, a different bench drew a stark "Lakshman Rekha,"  declaring with rigid finality that the power to modify was utterly non-existent under Section 34.

This created a jurisprudential black hole, forcing courts into an untenable choice between complete abdication—rubber-stamping an award with patent errors—and total annihilation—setting aside the entire award for a single, rectifiable flaw.

Beyond the Black Letter: Purposive Interpretation and Legislative Intent

Arguably, the most outspoken critique of Balaswamy is rooted in textualist fundamentalism. Sceptics posit that the Parliament's silence on modifying arbitral awards in the Act is an exclusive prohibition. This issue thus crystallises into a difficult doctrinal question: Can judicial inference fill a deliberate legislative omission?

This rhetoric, eloquently championed by Justice Viswanathan in his dissent, elevates the form over the substance and ignores the very purpose for which the 1996 Act was conceived. The Act's raison d'être is to promote an efficient, speedy, and fair dispute resolution mechanism. It is an affront to this core objective to suggest that a multi-million-rupee award, the product of years of tedious proceedings, must be entirely invalidated due to a patent mathematical error that could be rectified in a single court hearing.

We thus assert that the majority’s purposive interpretation does not usurp Parliament's role but rather breathes life into it.The legislative intent was to preclude a merits-based, appellate-style review, not to forbid the corrective intervention essential to cure a self-evident and outcome-altering blunder. Forcing parties back to square one imposes unnecessary hardship and financial burden upon the parties that chose this process for its cost-effective and expeditious nature, thereby defeating the statute's very purpose.

The dissent argues that courts cannot touch an arbitral award unless the statute gives them clear authority. On paper, this may sound faithful to legislative limits. But in practice, it traps the judiciary in procedural paralysis. Imagine an award with a simple interest miscalculation. The party must either accept the incorrect figure or spend months getting consent from all signed parties to reopen the arbitration just to correct a clerical error. The majority takes a more practical approach. It lets courts fix such obvious mistakes, not to change the law, but to make arbitration function as Parliament intended.

The Inevitable Corollary: How Severability Paved the Way for Modification

Furthermore, the ability to modify is not a new idea created out of nothing. It is a logical and unavoidable result of the established principle of severability, a principle that has deep roots in the Court's decisions under the 1996 Act.

This principle draws its vitality from the judicial landscape shaped by the landmark decision in ONGC Ltd v Saw Pipes Ltd, which empowered courts to review awards for 'patent illegality.' The logical consequence of finding such illegality in only one part of a multi-claim award was demonstrated in cases like J.G. Engineers (P) Ltd v Union of India, where the Court held that if a matter is severable, the court must segregate the award and set aside only the problematic part.

The majority took this established practice to its logical conclusion, stating unequivocally that "the limited and restricted power of severing an award implies a power of the court to vary or modify the award." The dissent’s attempt to distinguish between "severing" (permissible) and "modifying" (impermissible) is an exercise in semantic gymnastics that collapses under the weight of practical reality, a point extensively debated in legal commentary.

A Surgical Scalpel, not a Sledgehammer: Defining "Manifest Error"

The fear that permitting courts to correct “manifest errors” would surreptitiously introduce a merits review rests on two faulty premises – that judges cannot exercise restraint and that the threshold of intervention is low. We argue that both are misplaced apprehensions.

As a starting point, we need to remember the limitation that the court sets for modification: 

a. Courts can correct clear clerical or typographical errors, but only if this correction does not reopen the main issue and lead to a trial focused on the merits. 

b. If a problematic part of an award can be separated, only that part may be changed. 

c. Courts can change post-award interest only in exceptional cases, while pendente lite interest stays within the tribunal’s authority. 

d. In rare situations, the Supreme Court may use Article 142 to adjust an award to achieve complete justice.

This conjecture is not entirely new. Arbitral tribunals have already held similar power under Section 33 of the Act. The court’s role here does not involve interference; it focuses on efficiency. The goal is to prevent the delay and duplication that a formal remission under the statute would cause. History shows that courts have used their authority to correct arbitral awards responsibly. The ruling in J.C. Budharaja is one example where the Supreme Court reduced an award that exceeded the relief sought, effectively lowering it to the appropriate amount. Following the formal process for a remission in this case and gathering the consent of the parties to re-establish the tribunal would have been unnecessary.

Similarly, the benchmark for "patent illegality" is authoritatively laid by the Supreme Court in Associate Builders v Delhi Development Authority. It is not just any error of law, but something that "goes to the root of the matter." A manifest error is a species diluted within this genus of patent illegality. It is an error that is self-evident and requires no comprehensive submissions or re-appreciation of evidence. On realisation of this defect, the courts are already allowed to set aside in entirety or sever and enforce an award in part under s.34. We support the court on the conclusion that the presence of manifest error shall also be a legitimate ground for the limited modification of an award within the contours of s.34.

This is obviously not the same as rewriting the award due to unpalatable reasoning. Such intervention is liable to scrutiny as it crosses the lines s.34 and s.37 draw. Balaswamy, however, creates a functional tool that can be used only where sending the award back would be unnecessary or harmful. The line between a rectifiable error and a merit-based finding is clear, and the courts have identified and undisputably respected that line for decades.

To illustrate, if a contract specifies liquidated damages at ₹1 lakh per day and the arbitrator correctly finds a 20-day delay but calculates the damages as ₹2 lakh instead of ₹20 lakh, that is a manifest computational error ripe for modification. However, if the arbitrator, after weighing evidence, determines the delay was only 10 days, not 20, that finding on the merits is sacrosanct and beyond the court's modifying power. This distinction is the bedrock of the majority’s ruling, a surgical scalpel designed to excise a cancerous error, not a sledgehammer to demolish the entire edifice.

A Pragmatic Distinction: The Power to Modify Post-Award Interest

Nowhere is the majority’s pragmatic approach more evident than in its nuanced handling of interest modification, a point often seized upon by critics as the prime example of judicial overreach. This criticism, however, ignores the crucial firewall the Court has erected between different periods of interest, thereby safeguarding the arbitrator's core domain. The judgment draws a bright line between pendente lite interest (from cause of action to the award) and post-award interest (from award to payment).

The Court holds that pendente lite interest, being a matter of the arbitrator’s discretion based on the merits and evidence presented, cannot be modified by a court under Section 34. If found to be patently illegal (e.g., contravening an express contractual bar), the court’s only power is to set it aside, not to substitute with its own rate. This finds resonance in the judicial self-restraint shown in cases like Krishna Bhagya Jala Nigam Ltd v G Harischandra Reddy, where the Supreme Court, to modify an interest rate, had to invoke its extraordinary powers under Article 142, implicitly acknowledging that no such general power vests in courts hearing a Section 34 petition.

The true innovation lies in the court's handling of post-award interest, which is not an adjudication on past events but a provision for future compliance. The majority rightly recognized that an arbitral tribunal is not clairvoyant; it cannot foresee extensive delays during prolonged challenge proceedings under Sections 34 and 37. An interest rate that is compensatory at the time of the award can become grossly punitive over time due to supervening circumstances like a drastic fall in commercial lending rates. This is not a hypothetical fear; in Vedanta Ltd v Shenzden Shandong Nuclear Power Construction Co Ltd, the Supreme Court itself modified post-award interest to reflect commercial realities and prevent an unjust windfall.

The power to modify post-award interest, therefore, is not about second-guessing the arbitrator’s finding but about ensuring the remedy remains equitable in light of post-award events. A court cannot modify this interest simply because it disagrees with the rate; the modification must be justified by circumstances arising after the award that render the original rate so unconscionable as to shock the conscience of the court and thus become patently illegal in its effect. This carefully calibrated power is a necessary tool for justice, not a license for interference, ensuring that the fruits of arbitration are not poisoned by the passage of time.

Debunking the Myth of International Unenforceability

Finally, the concern regarding the enforceability of a modified award under the New York Convention is the ultimate red herring; the primary objection raised was that a modified award would "merge" with the court's order, becoming a "court decree" and thus unenforceable under the Convention, which applies only to "arbitral awards."^[1]

Firstly, the idea of courts stepping in to make limited corrections to arbitral awards is well-established internationally. Other Model Law jurisdictions allow similar, limited court variation. Section 49(8)(b) of Singapore’s Arbitration Act 2001 lets courts vary awards where a question of law substantially affects the parties’ rights and the tribunal’s decision is “obviously wrong”, “raises serious doubt”, or where court intervention is “just and proper”. The Singapore Academy of Law’s 2020 report even urged extending that approach to international cases by amending the International Arbitration Act 1994. Australia follows a similar path: the Commercial Arbitration Act 1986 contains parallel modification provisions in sections 38(3)(a) and 38(7).

Secondly, the claim that an internationally enforceable award becomes vulnerable once a court makes limited modifications also does not hold. The Court dismantled this by noting that the doctrine of merger does not apply to s.34 proceedings, which are not appellate in nature but are for setting aside an award. More crucially, the Court focused on the precise language of Article V(1)(e) of the Convention, which allows refusal of enforcement if an award "has not yet become binding on the parties" under the law of the seat. The Court reasoned that an award, once modified by the supervisory court at the seat, is the only version that is legally "binding" in India. Therefore, for international enforcement, the award "as modified by the judgment/order under Section 34" is the final, binding arbitral award.

The Court correctly concluded that to hold otherwise would create an absurd situation where an award found to be partially illegal by the seat court would still have to be enforced abroad in its original, flawed form.

Much Ado About Nothing

The majority opinion in Gayatri Balaswamy is a deliberate shift away from a rigid, binary interpretation of the scope of powers under Section 34. Though concerns about statutory limitations hold are persuasive, these reasons are compelling enough to believe that narrowly circumscribed powers of modification can align with the broader structure of the 1996 Act, while also advancing its raison d'être of expeditious dispute resolution.

Post-1991 liberalisation, arbitration was rebranded to appeal to international commercial players. These users prize predictability: in both process and outcome. That’s why they prefer arbitration over traditional courts.

The present ruling helps in clarifying on how the courts may tweak awards. Courts aren’t outsiders to arbitration; they’re integral to its architecture. Their job is to ensure the system works smoothly, not to bounce parties back to square one. By limiting the scope for correction, the court has improved predictability. Now, parties know exactly what errors can be addressed during an appeal.

How these new judicial powers play out indeed remains to be settled through future rulings; however, an examination of existing jurisprudence suggests that these developments do not necessarily risk branding India as hostile to arbitration.

India needs carefully calibrated judicial oversight that remedies chronic challenges plaguing its arbitration landscape. Although it is really unlikely that the tension between judicial oversight and the principle of arbitral finality will be resolved once and for all, the efforts to recalibrate this tussle into a workable equilibrium are nevertheless praiseworthy.

---

^[1] First Year BA LLB (Hons) Student at Gandhinagar National Law University [E-mail: prabhaskumar2607@gmail.com]

^[2] Second Year BA LLB (Hons) Student at National Law University Odisha [E-mail: 24ba097@nluo.ac.in]

^[1] Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York, 10 June 1958, entered into force 7 June 1959) 330 UNTS 38 (New York Convention).