The Devil in the Proviso? India's Section 36(3) Fraud Exception and Award Enforcement

The subject of automatic stay of arbitral awards in India, arising from the Supreme Court of India’s (the “Supreme Court”) interpretation of Section 36 of the Arbitration and Conciliation Act, 1996 (“Act”), has long troubled India’s pro-arbitration turn. In National Aluminium Co. Ltd. v. Pressteel & Fabrications Pvt. Ltd. (“NALCO”), the Supreme Court read Section 36, which governs enforcement of arbitral awards under the Act, to mean that once a party filed a Section 34 challenge, namely an application to set aside an award, the award could not be enforced until the challenge was decided. This result was that an award-holder, despite having completed the arbitral proceedings, could still be denied the fruits of the award merely because the losing party had initiated setting-aside proceedings.

The 2015 Amendment sought to undo this mischief by clarifying that filing a Section 34 application would not,  on its own,  stay enforcement. In BCCI v. Kochi Cricket Pvt. Ltd. (“BCCI v. Kochi”), Justice R.F. Nariman, writing for the Supreme Court, held that amended Section 36 would apply even to pending Section 34 applications, since no vested right inhered in an award-debtor to resist enforcement through automatic stay. Thereafter, in Hindustan Construction Company v. Union of India (“Hindustan Construction”), the Supreme Court struck down Section 87 of the Act, which had attempted to revive the pre-2015 position. Prior commentary has traced this movement through the aftermath of BCCI v. Kochi, the Supreme Court’s invalidation of Section 87 in Hindustan Construction, and India’s broader pro-enforcement trajectory after the 2015 amendments.

The 2021 Amendment created a “wolf in sheep’s clothing” (discussed here) problem in India’s award-enforcement regime: a proviso ostensibly introduced to preserve arbitral integrity by addressing fraud and corruption may, in effect, reintroduce a serious hurdle to enforcement by mandating an unconditional stay once its threshold is met. The second proviso to Section 36(3) provides that where the court is satisfied that a prima facie case exists that the arbitration agreement, the underlying contract, or the making of the award was induced or affected by fraud or corruption, it shall stay the award unconditionally pending disposal of the Section 34 challenge. A later review of Indian arbitration developments in 2021 (critiqued here) similarly noted that the amendment had generated apprehension that recalcitrant parties may again deprive award-holders of the fruits of their awards by invoking fraud or corruption. The controversy, therefore, is not whether awards infected by fraud or corruption deserve protection from enforcement; rather, it is whether a prima facie finding should produce an unconditional stay without allowing courts to tailor relief to the circumstances.

The issue has acquired renewed relevance in light of the Draft Arbitration and Conciliation (Amendment) Bill, 2024, which reflects India’s continuing attempt to recalibrate its arbitration regime. However, while the proposed reforms seek to strengthen the institutional and procedural landscape of arbitration, they do not squarely address the remedial difficulty created by the second proviso to Section 36(3): whether a preliminary finding of prima facie fraud or corruption should automatically result in an unconditional stay of enforcement.

The Conflict

The Supreme Court’s movement from NALCO to BCCI v. Kochi and Hindustan Construction rested on a simple premise: arbitral finality must have practical effect. Section 36, post-2015, was designed to ensure that a losing party could not convert a Section 34 challenge into an automatic stay on enforcement. The award-debtor must specifically seek a stay; the court must decide whether a stay is warranted; and, ordinarily, such a stay may be conditioned on deposit, security, or other safeguards.

The second proviso to Section 36(3), however, sits uneasily with that design. On the one hand, the Act rejects the proposition that enforcement should be suspended merely because an award is under challenge. On the other hand, the 2021 Amendment commands an unconditional stay where fraud or corruption prima facie affected the arbitration agreement, the contract, or the making of the award. The conflict is therefore not between enforcement and integrity simpliciter. It is between two remedial instincts within the Act: one resisting delay by award-debtors, and the other protecting arbitration from awards tainted by fraud or corruption.

Both instincts are defensible. fraud and corruption strike at arbitral legitimacy, and no pro-enforcement policy can require courts to shut their eyes to serious allegations that the process itself has been compromised. However, the proviso resolves this tension through an unusually rigid consequence. Once the prima facie threshold is crossed, the court “shall” stay the award unconditionally. It is this movement from suspicion to suspension, without any intermediate safeguard, that creates the central difficulty.

Recent authority brings this difficulty into sharper relief. In Union of India v. Rashmi Metaliks Ltd. (“Rashmi Metaliks”), the Orissa High Court held that the second proviso is attracted only where the alleged fraud or corruption is of a serious and fundamental nature, bears a direct nexus with the relief granted, and is supported by credible material rather than speculation. Yet the Orissa High Court also recognised that once such prima facie satisfaction is reached, an unconditional stay becomes mandatory. Rashmi Metaliks, therefore, captures the proviso’s two-sided character: the threshold may be high, but the consequence, once triggered, leaves no room for remedial calibration.

The Fraud Exception

1.     Threshold and Consequence

The second proviso is framed as a distinct integrity-preserving measure. In principle, this is unobjectionable. Arbitration derives its authority from consent, procedural fairness, party autonomy, and finality. Where the arbitration agreement, the underlying contract, or the making of the award is affected by fraud or corruption, the case for immediate enforcement weakens. However, the vice lies in the relationship between threshold and consequence. A preliminary finding does not finally establish fraud or corruption, nor does it determine whether the award will ultimately be set aside under Section 34. Yet that preliminary satisfaction produces an unconditional stay until the disposal of setting-aside proceedings.

In commercial terms, this may be far from provisional. A stay granted at the threshold may operate for years, during which the award-holder is deprived not only of payment, but also of the leverage and finality that a favourable award is meant to confer. The stay is interim in form, but may be final in effect. That is why the proviso is not merely a delay problem; it is a remedial design problem.

2.     Pleading Arbitrage

The proviso also creates pleading arbitrage. If an ordinary Section 34 challenge, based on limiting setting-aside grounds such as patent illegality, perversity, or public policy leads only to discretionary stay, often accompanied by deposit or security, while fraud or corruption opens the door to unconditional stay, award-debtors may, from the outset, draft their objections in the vocabulary of fraud. The risk is not confined to plainly unfounded allegations. Contractual interpretation disputes, evidentiary objections, procedural complaints, allegations of bias, or disagreements with the tribunal’s appreciation of facts may be dressed up as fraud or corruption to obtain a stronger remedial consequence. In West Bengal Industrial Development Corporation Ltd. v. Tata Motors Ltd., the Calcutta High Court’s refusal to treat bias as a ground for unconditional stay under the second proviso is therefore significant, for it polices the boundary between genuine fraud and collateral attacks on the tribunal.

This boundary matters because Section 34 scrutiny is meant to be narrow. Some authors, analysing the Court’s approach to scrutiny of awards, have noted that post-2015 review under Section 34 is not an appellate exercise, and that public policy is confined to limited categories, including fraud or corruption in the making of the award. Other commentary on public policy in international commercial arbitration has similarly warned, in discussions on the limits of public policy review and the narrow scrutiny of awards under Section 34, that such objections must not become a disguised channel for merits review. By the same logic, the stay stage under Section 36 should not become a shadow Section 34 hearing, where courts are invited to conduct merits review under the language of fraud.

A narrower construction of the proviso would therefore require courts to separate bona fide fraud that truly affects the arbitration agreement, the contract, or the making of the award, from allegations that fall outside the statutory purpose of the proviso, merely attack the tribunal’s reasoning or evidentiary conclusions. If the alleged fraud requires extensive factual inquiry, cross-examination, or reappraisal of evidence, it should ordinarily be rejected at the threshold for the purpose of unconditional stay and determined, if necessary, in the setting-aside proceedings.

3.     The Missing Middle Path

The absence of a remedial middle path further sharpens the problem. In Popular Caterers v. Ameet Mehta, the Supreme Court considered unconditional stay outside the second proviso, but treated such relief as exceptional rather than routine. Similarly, in International Seaport Dredging Pvt. Ltd. v. Kamarajar Port Ltd., the Supreme Court modified a stay order by directing deposit of 75% of the awarded amount, reflecting the ordinary Section 36(3) logic of security as a middle path. The contrast is instructive. Article VI of the New York Convention permits a court, where an application for setting aside or suspension has been made at the seat, to adjourn enforcement and, where appropriate, order suitable security. This is a discretionary and calibrated model: the court may adjourn enforcement, and may also require security. By contrast, the second proviso to Section 36(3) adopts a mandatory model, requiring an unconditional stay once its threshold is met. The point is not that fraud or corruption should be treated lightly. Rather, the law can take such allegations seriously without making unconditional stay the only arrow in the court’s quiver.

Conclusion

The second proviso to Section 36(3) addresses a real concern. Arbitration cannot command legitimacy if courts are compelled to enforce awards seriously tainted by fraud or corruption. Yet legitimacy also suffers when fraud becomes a procedural shortcut to delay enforcement, and in practical terms defeats enforcement. Indian courts should therefore treat the proviso as exceptional. A prima facie case should require strong contemporaneous material, and not merely skilful pleading.

Legislative clarification may also be desirable. A better formulation would allow courts to stay enforcement in serious fraud or corruption cases while preserving discretion to impose security, deposits, or other conditions. This could be achieved by replacing the mandatory formulation “shall stay the award unconditionally” with language such as “may stay enforcement subject to such conditions as the court considers appropriate”. Such an approach would protect both sides of the arbitral bargain: the integrity of the process and the enforceability of the award.

Fraud must remain a ground for resisting enforcement in serious cases. It should not become the vocabulary through which automatic stay is revived in substance, though not in name.