This article has been published by Supreme Court Observer at: Modification of arbitral awards: a practitioner’s perspective – Supreme Court Observer
The recent Constitution Bench decision reaffirms speed and economy but it has a flipside that could threaten the finality principle.
On 30 April, a five-judge Constitution Bench delivered a landmark judgement in Gayatri Balasamy v ISG Novasoft Technologies. In a 4:1 majority, the Bench upheld a court’s power to modify an arbitral award under Sections 34 and 37 of the Arbitration Act, 1996.
The law now settled
The majority Judgement authored by Chief Justice Sanjiv Khanna held that Section 34 of the Act, which empowers courts to set aside an award, also confers the power to modify an award in “limited” circumstances. It found that the power of a court to set aside an award also includes the power to partially set it aside. It cautioned that this power cannot be exercised if the valid and invalid portions of the award are interdependent or intrinsically intertwined.
The majority further held that the Act’s “silence” on the power to modify an award does not amount to a prohibition. In their view, this power is inherent as it is incidental to the power to set aside—this implied power is necessary to ensure that the object of the Act is advanced and hardship is avoided.
Significantly, the majority highlighted that the modification of an award cannot be placed on the same pedestal as an appeal. This is because a Section 34 court can modify an award only if there is no uncertainty regarding the intended modification. The majority described the limited situations under which an award may be modified:
- if the denial to modify an award would impose significant hardships, escalate costs and lead to unnecessary delays;
- if the alternative is to force parties into a new arbitration only to “…affirm a decision that could easily be arrived at by the court.”
Additionally, the majority upheld the court’s power to modify post-award interest under Section 31(7) of the Act. This provision is a unique creation under Indian law and was not derived from the UNCITRAL Model Law. The Bench reasoned that post-award interest can be scrutinised by courts against the standards set in Section 31(7)(b).
Moreover, the majority held that the power to remand an award is discretionary. In fact, it noted that if there are serious acts of omission, commission, substantial injustice or patent illegality or if remand does not serve parties’ interests in time-sensitive matters or where it would lead to undue costs and inefficiencies, the power to remand ought not to be exercised.
Finally, the majority noted that the Supreme Court also has the power to modify an award under Article 142 of the Constitution. However, they cautioned that the Court cannot rewrite an award using this power.
Justice K.V. Viswanathan’s sole dissent rejected the proposition that a court has the power to modify an award under Section 34. He held that reading in the power to modify is antithetical to the objective of the Act to minimise judicial intervention. He also observed that it can negatively impact the enforcement of awards under the New York Convention on the Recognition and Enforcement Foreign Awards, 1958.
Guardrails for modification
By allowing only the “flawed” portion that is indisputably severable to be modified, the Judgement attempts to draw a tight boundary around the power. It also notes that modification ought to be carried out when it is the only way to avoid injustice, delay or excessive costs.
In practice, however, the boundaries may be more ambiguous because the Supreme Court hasn’t defined the extent to which a court can look into the merits of the case while assessing whether to modify, set aside or remand. Phrases like “furthering the ends of justice” (Paragraph 40) and “error not apparent on the face of the award” (Paragraph 54) are open‐ended and merely substitute one form of uncertainty for another.
Further, there is no clear test to distinguish valid clauses from invalid ones, and to judge what counts as “hardship” or “unnecessary delay”. Practitioners have long sought a bright line on these issues against a backdrop of conflicting precedents. Such a contour is not forthcoming in this Judgement.
Also—the majority does not identify situations where a court should lean towards remand instead of modification or setting aside. If the test is to see which process is more expeditious, one may argue that modification would be the preferred method. A remand would lead to an additional round of contentions before the tribunal, whose decision would once again be subject to judicial scrutiny under Section 34. While one can expect that courts will lean in favour of remand where extensive reconsideration of fact or merits is required, it will be interesting to see how courts navigate this distinction in the future.
Not consistent with international practice
The majority’s claim that its views are consistent with international practice may be flawed. Admittedly, some leading arbitration jurisdictions grant courts limited powers to modify awards. However, these powers are expressly set out in statute.
UK and Singapore, for example, are two arbitration-friendly jurisdictions where there are strict guardrails about the circumstances in which an award can be judicially challenged. However, the arbitration statutes of both countries contain express provisions granting the courts the power to modify an award on limited grounds.
From a practice perspective, an express statutory power provides more clarity and limitations than what can be discerned from the Judgement.
India’s latest Arbitration and Conciliation Bill (2024) does not expressly provide any power to modify an award. In the wake of this Judgement, it will be interesting to see if the Bill is updated to include the recommendations on modification from the T.K. Viswanathan Expert Committee Report.
Enforcement under the NY Convention
Under the New York Convention, an arbitral award need not be confirmed by the seat court before it is enforced in another jurisdiction. Under Indian law, challenging an award under Section 34 does not automatically stay its enforcement. This often leads to a situation where the successful party attempts to enforce the award in a foreign jurisdiction under the New York Convention, while challenge proceedings are still pending in India.
This issue has been tackled by the English Arbitration Act to some extent, which provides that a modified award becomes part of the original award. There is no such provision in the Indian Act. If an Indian court modifies an award under Section 34, there is likely to be a discrepancy between the original award enforced abroad and the altered version surviving in India. The Judgement does not deal adequately with this conundrum.
The path ahead
By expressly mentioning cost-cutting and deadline‑driven justice, Gayatri Balasamy attempts to reaffirm arbitration’s core promise: economy and speed. Its endorsement of a narrow “modification” power is not just a concession; it is a targeted fix to ensure full justice without dragging parties through endless rounds of litigation or fresh arbitration.
We foresee this Judgement to be of assistance to awards where the alternative to modification is the continuation of a protracted ad-hoc arbitration. The Judgement may also incentivise parties to adopt arbitration with the knowledge that a supervisory court has the power to modify an award to protect their interests.
The flipside, however, is that the ruling could add to the prevailing trend of Section 34 challenges by creating a fresh opportunity to litigate. Additionally, its findings on the modification of interest may encourage courts to modify awards where the tribunal has expressly not granted any interest.
Ultimately, the real marker of assessment will be how courts exercise the power to vary going forward. In our view, we may not have seen the last of Gayatri Balasamy and the law it lays down.