Introduction
Arbitration promises parties a private, efficient, and largely final dispute resolution process. Indian courts have long debated how far they may touch an award. Can a court modify an arbitral award, or only set it aside? This question troubled litigants and lawyers for nearly three decades. The Supreme Court finally answered it in Gayatri Balasamy v. ISG Novasoft Technologies Ltd. A five judge Constitution Bench recognised a narrow power of modification in 2025. Understanding this shift matters for every commercial party choosing arbitration in India. It shapes how contracts get drafted and how disputes eventually get resolved.
The Statutory Framework Under Section 34
The Arbitration and Conciliation Act, 1996, governs domestic arbitration in India. Section 34 allows a party to apply for setting aside an award. The provision lists specific grounds, including incapacity, invalid agreement, and procedural unfairness. Notably, the section never uses the word modify anywhere in its text. Parliament modelled this provision closely on the UNCITRAL Model Law of 1985. That Model Law deliberately excludes any power of substantive modification for courts. This choice reflected a global consensus favouring minimal judicial interference in arbitration. Drafters believed that broad curial power would discourage parties from choosing arbitration. They wanted arbitration to remain distinct from ordinary, appeal heavy civil litigation.
Indian legislators largely accepted this reasoning when enacting the 1996 Act. Indian courts, however, developed inconsistent practices despite this seemingly clear legislative intent. Some High Courts modified awards routinely, treating this as an implied power. Others read Section 34 strictly, permitting only annulment of a flawed award. This split created genuine uncertainty for parties challenging or defending arbitral awards. Litigants often could not predict which approach a particular bench would follow. Such unpredictability undermines the very efficiency that arbitration is meant to offer. Commercial parties need confidence that court intervention will follow settled, consistent principles.
The Road to Gayatri Balasamy
The conflict became especially visible after Project Director, NHAI v. M. Hakeem in 2021. A two judge bench held that Section 34 permits no modification whatsoever. It reasoned that courts exercising this power act only in a supervisory role. Many arbitration practitioners criticised this ruling as excessively rigid and impractical. They argued that minor, severable errors should not require fresh arbitration proceedings. The Gayatri Balasamy dispute itself arose from a workplace harassment and termination case. An arbitral tribunal awarded compensation, which a single judge later enhanced substantially. The Division Bench reduced this enhancement sharply on appeal under Section 37.
Both parties approached the Supreme Court, prompting a reference to a larger bench. This reference asked whether M. Hakeem had correctly interpreted Section 34’s true scope. The reference bench itself noted the significant practical hardship caused by strict annulment. It observed that forcing fresh arbitration for minor errors, wastes considerable time and money. This observation ultimately shaped how the Constitution Bench approached the wider question.
The New Doctrine of Limited Modification
On 30 April 2025, the Constitution Bench delivered its answer by majority. Chief Justice Sanjiv Khanna held that courts possess a limited modification power. He identified four specific situations where such modification remains constitutionally permissible. These include severing an invalid portion from an otherwise valid award. They also include correcting clerical, typographical, or computational errors within an award. Courts may further adjust the rate of post award interest where appropriate.
Finally, the Supreme Court retains power to modify awards using Article 142. Justice K.V. Viswanathan dissented sharply, insisting Section 34 permits no modification at all. He warned that this new doctrine blurs arbitration with ordinary civil appeals. The majority, however, viewed this limited power as necessary for practical justice. Justices Gavai, Kumar, and Masih joined the Chief Justice in this reasoning. They emphasised that finality should not become an excuse for unnecessary delay. Together, they formed a clear four judge majority on this central question.
Drawing the Line Between Modification and Merits Review
The Court’s four categories attempt a careful doctrinal balancing act. Severability lets courts separate valid findings from an isolated legal error. This differs sharply from courts reassessing evidence or substituting their own findings. Clerical correction similarly touches form, not the substance of an arbitral decision. Interest rate adjustment, too, addresses a mechanical consequence rather than the core dispute. Article 142, meanwhile, remains reserved for truly exceptional circumstances before the Supreme Court. Legal scholars nonetheless warn that these categories lack precise, self enforcing boundaries.
A court correcting a clerical error might easily drift toward reassessing merits. Future litigation will likely test exactly how far these categories can stretch. Lower courts must now develop consistent practice within this newly drawn framework. Uniform application across High Courts will take time and repeated appellate clarification. Until then, some inconsistency in applying these categories seems almost inevitable. Practitioners should therefore watch closely how different High Courts interpret this new doctrine.
Comparative Perspectives On Curial Intervention
England’s Arbitration Act, 1996, expressly allows limited appeals on points of law. Section 69 permits this only where parties have not excluded such appeals. Singapore’s International Arbitration Act largely mirrors the Model Law’s restrictive approach instead. Australian courts have occasionally read implied modification powers into their arbitration statutes. These comparative examples show no single global consensus on modification powers. Legal systems balance finality and fairness differently based on local judicial culture. India’s new four category approach sits between these more extreme comparative positions. It preserves broad finality while permitting correction of narrow, defined defects.
This middle path reflects India’s evolving identity as an arbitration friendly jurisdiction. Foreign investors and businesses increasingly consider such doctrinal clarity when choosing India as a seat. India has also amended its arbitration law repeatedly to attract international commercial disputes. This latest judicial clarification adds further predictability to that ongoing legislative effort. Countries competing to become preferred arbitration seats watch these developments closely and carefully. India’s ranking in global arbitration surveys may improve as a result of this clarity.
Impact On Stakeholders and Future Practice
Commercial parties will now negotiate contracts with this modification doctrine firmly in mind. Drafting lawyers must anticipate which category a future challenge might realistically fall under. Arbitral institutions may also revise their procedural rules to minimise correctable errors upfront. Tribunals themselves face greater pressure to issue precise, well reasoned, and error free awards. Insurance companies and large corporations, frequent users of arbitration, welcome this added certainty. Smaller businesses and individual litigants, however, may still struggle with unpredictable litigation costs. Fresh arbitration, even where avoided under this doctrine, remains expensive for less resourced parties. Legal aid and pro bono support could help bridge this persistent access to justice gap. Civil society organisations could also offer simplified guidance for individuals navigating arbitration disputes.
Suggestions
Parliament should consider codifying these four categories through formal legislative amendment. This would remove ambiguity currently left to judicial interpretation alone. Courts should develop clear guidelines distinguishing genuine correction from disguised merits review. Arbitral institutions should train tribunals to minimise clerical and computational errors upfront. This would reduce the need for later curial intervention altogether. Legal education must update arbitration curricula to reflect this significant doctrinal shift. Practitioners should draft arbitration clauses anticipating this now clarified scope of court power. Bar associations could publish practice notes explaining these four categories in simple terms.
Judicial academies should include this doctrine in ongoing training for commercial court judges. Law schools ought to update arbitration syllabi to reflect this landmark ruling promptly. Legal journals and bar councils should encourage continued scholarly debate on unresolved boundary questions. Policymakers might also study whether a statutory cap on modification scope proves useful. Such a cap could prevent future benches from expanding these categories too broadly. A future amendment could also formally overrule M. Hakeem to remove any lingering doubt.
Conclusion
Gayatri Balasamy v. ISG Novasoft Technologies Ltd. closes a long standing interpretive debate. It replaces rigid textualism with a carefully limited, functional approach to justice. The judgment protects arbitral finality while allowing correction of genuine, narrow defects. Its lasting success will depend on how consistently courts apply these categories. If applied faithfully, it could strengthen confidence in Indian arbitration significantly. If stretched too far, it risks reviving the very unpredictability the Constitution Bench sought to end.
References
Gayatri Balasamy v. M/S ISG Novasoft Technologies Limited, 2025 INSC 605.
Project Director, NHAI v. M. Hakeem, (2021) 9 SCC 1.
The Arbitration and Conciliation Act, 1996, Section 34.
UNCITRAL Model Law on International Commercial Arbitration, 1985.
Arbitration Act, 1996 (United Kingdom).
