The Illusion of Restraint: The Paradox of Supreme Court Jurisprudence under Article 226 and 227 in Arbitration
Introduction
Arbitration in India was envisioned as a self-contained dispute resolution mechanism, free from any excessive judicial intervention. This idea is embedded in Section 5 of the Arbitration and Conciliation Act, 1996, (‘the Act’) which makes it clear that courts should step into arbitral proceedings only in situations expressly permitted by the statute.
Yet, in practice, this ideal has proven difficult to maintain. Parties frequently approach High Courts under Articles 226 and 227 of the Constitution to challenge arbitral orders and proceedings, sometimes even for the smallest inconveniences at the very outset of arbitration. This has prompted the Supreme Court to respond with a line of jurisprudence that seeks to restrict this routine access by litigants. Time and again, it has been emphasized that writ jurisdiction should be exercised only in cases of the rarest nature, where the error is so fundamental, or the lack of jurisdiction is so obvious that it cannot be ignored. At the same time, the Court has been reiterating that while statutory schemes like the Act seek to limit judicial intervention, they cannot extinguish the High Courts’ constitutional powers.
However, the difficulty lies in the nature of this restrictive standard itself. The thresholds invoked by the Court – “rarest cases,” or “manifest injustice” – are articulated in broad terms, offering little objective guidance as to when intervention is truly warranted. This creates interpretive space within which litigants are able to recharacterize routine procedural grievances as jurisdictional defects. The result is that the so-called restrictive line ends up enabling intervention to much the same extent as a permissive one.
This blog engages with this structural paradox in the Court’s approach to writ intervention. It begins by exploring the statutory scheme of the Act, and the constitutional framework within which it operates. It then examines the two parallel lines of Supreme Court jurisprudence: the restrictive line and the permissive line. The analysis shows that, despite its language of caution, the restrictive line of cases ends up enabling judicial interference to much the same extent as the permissive one. It argues that the imprecise language underlying the restrictive standard creates a fundamental contradiction that weakens the arbitral process itself. It concludes by examining how this ambiguity plays out in real disputes and by suggesting a clearer, more principled standard for reconciling the Act’s aim of minimal judicial intervention with constitutional powers.
The Statutory Framework and its Constitution Tension
In order to reduce judicial intervention, the Act is designed as a self-contained code. This goal is embodied in Section 5, which states that judicial intervention in arbitral disputes is prohibited unless specifically allowed by the Act.
This design is reinforced by the Act’s internal architecture. Section 16 incorporates the principle of kompetenz-kompetenz, which gives the arbitral tribunal the power to rule on its own jurisdiction and mandates that jurisdictional objections be brought up and resolved during the arbitral process. While Section 37 restricts appellate interference to a specific category of orders, Section 34 offers a narrowly defined mechanism for overturning arbitral awards. When combined, these clauses establish a closed procedural framework designed to guarantee that judicial intervention occurs only after the arbitral process has concluded. In SBP & Co. v. Patel Engineering, the Supreme Court recognized this structure, describing the Act as a “complete code,” stressing that parties must operate within it rather than pursue collateral judicial review.
This intention is further supported by the legislative history. The 246th Law Commission report noted that excessive judicial intervention has traditionally hampered the growth of arbitration in India and emphasized the need to limit such intervention. India’s adoption of the UNCITRAL Model Law further supports this intent: Article 5 permits judicial involvement only where the law expressly provides it.
This statutory framework however functions alongside the constitutional scheme. Articles 226 and 227 confer broad supervisory powers on High Courts, which have been continuously held to be unaffected by ordinary legislations. Consequently, notwithstanding the Act’s emphasis on minimal judicial intervention, writ jurisdiction continues to be invoked to challenge orders mid-proceedings, often on grounds of lack of jurisdiction or violation of principles of natural justice – as the subsequent sections will illustrate through specific instances of such challenges.
The Two Lines of Supreme Court Jurisprudence
Judicial engagement with arbitration-related writ petitions has developed along two distinct lines. The first is a restrictive approach that emphasizes the need to preserve the statutory framework of the Act. The second is a permissive approach that reaffirms the availability of constitutional remedies despite the presence of an arbitral mechanism. Both lines continue to operate concurrently and are frequently cited before High Courts.
The Restrictive Approach
The restrictive line begins with the premise that arbitration should not be interrupted except in clearly delineated circumstances. In Premjibhai K. Shah v. Sardar Sarovar Narmada Nigam Ltd., the Supreme Court stated that recourse to Article 226 or 227 in arbitration matters should arise only in “exceptional rarity.” The Court observed that the Act provides a complete set of remedies and that judicial intervention outside these statutory pathways should be avoided.
In Deep Industries Ltd. v. ONGC, the Court clarified that interference with interlocutory orders under Article 226 or 227 may disrupt the arbitral process and should be undertaken only when an order reflects a “patent lack of jurisdiction.” The Court warned that entertaining such petitions routinely would undermine the efficiency of arbitration.
Similarly, in Punjab State Power Corporation Ltd. v. EMTA Coal Ltd., the Court held that intervention is permissible only when an error is so evident that it “stares one in the face,” indicating an extremely narrow threshold. The Court discouraged judicial review of Section 16 rulings prior to the conclusion of the arbitral proceedings.
In Bhaven Construction v. SSNNL, the Court reaffirmed that constitutional remedies cannot be excluded but emphasized that High Courts should refrain from exercising them when the statute provides an internal mechanism. The Court linked intervention to circumstances where a party is left without any remedy or where the proceedings appear to involve ‘bad faith.’
Collectively, these decisions indicate the Court’s position that writ jurisdiction should not function as an alternative appellate mechanism in arbitration matters and that parties are expected to adhere to statutory remedies.
The Permissive Approach
Alongside this restrictive view, the Court has issued decisions acknowledging that constitutional jurisdiction remains available despite the presence of an arbitral mechanism. In Uttar Pradesh Power Transmission Corporation Ltd. v. CG Power and Industrial Solutions Ltd., the Supreme Court held that the existence of an arbitration clause does not bar the High Court from exercising writ jurisdiction where appropriate grounds are made out.
In Unitech Ltd. v. Telangana State Industrial Infrastructure Corporation, the Court held that contractual disputes involving the State may still raise issues of public law, and the presence of an arbitration clause does not prevent judicial review when State action is alleged to violate constitutional standards.
In Union of India & Ors. v. Tantia Construction Pvt. Ltd, the Court reiterated that the availability of an alternative remedy, including arbitration, does not preclude the High Court from exercising jurisdiction under Article 226. The Court described constitutional review as a remedy that cannot be ousted by statute.
These decisions emphasize that constitutional remedies remain accessible and that their availability does not depend on the contours of the arbitral framework. While the permissive line confirms that constitutional jurisdiction remains open notwithstanding the arbitral framework, it does not itself create the central doctrinal difficulty examined in this paper. The existence of permissive judgments is unsurprising given the constitutional basis of Articles 226 and 227. The more significant issue arises from the restrictive line, which is intended to limit access to writ remedies but relies on standards that lack precision and have not produced consistent or predictable outcomes. The analysis that follows therefore focuses primarily on the restrictive approach and the ways in which its indeterminate thresholds contribute to continued judicial intervention in arbitration matters.
The Limits of the Restrictive Jurisprudence
The Restrictive line of jurisprudence formally casts Articles 226 and 227 as remedies of narrow and exceptional reach. Courts repeatedly emphasize that interference is warranted only where perversity “stares in the face,” where there is a patent lack of jurisdiction, or where manifest injustice is demonstrable. On paper, this language signals a high threshold designed to discourage mid-stream constitutional challenges.
In practice, however, litigants continue to access writ courts during ongoing arbitrations by recasting routine procedural objections within this exceptional vocabulary. Recent decisions demonstrate how challenges to Section 16 determinations, amendment of pleadings, framing of issues, evidentiary rulings, and cross-examination timelines are routinely styled as violations of natural justice and thus brought under the ambit of this “strict” threshold.
In Oriel Financial Solutions v Bestech Advisors, the rejection of a Section 16 plea was brought under Article 227 and adjudicated through the “shock the conscience” lens before the Court ultimately declined interference. In AMR-BBB Consortium v Bharat Coking Coal & Shahnaz Hussain v Suresh Dhanuka, quintessential case-management decisions were similarly constitutionalized and subjected to supervisory scrutiny despite repeated reminders of restraint by the judiciary. The same pattern is visible in Aneja Constructions v Doosan Power Systems & Lata Yadav v Shivakriti Agro, where disputes concerning pleadings timelines and termination applications were framed as exceptional circumstances warranting intervention. Even more revealing is Dexter Capital Advisors v. Serosoft Solutions, where the High Court, while acknowledging that interference is ordinarily “least warranted” still granted additional time or cross-examination on natural justice grounds, an intervention later set aside by the Supreme Court. In MDD Medical Systems v. DIAC, the Delhi High Court was asked to interfere with an MSME arbitral tribunal’s decision to revive proceedings after procedural closure. While the Court ultimately deferred to Section 16 of the Act, it still entertained the petition and conducted a merits hearing. Likewise, in Lalit Mohan v. NAFED, a litigant sought to restrain the arbitrator from issuing the final award by invoking IBC moratorium concerns. These cases collectively show that the restrictive formulation does not prevent petitions from being brought and heard.
The cumulative effect is to dilute the deterrent force of the restrictive standard itself. The institutional cost to arbitration arises not only when interference is granted, but when writ courts are drawn into mid-proceeding supervision through a merits review. The indeterminate expressions that anchor the doctrine provide sufficient argumentative elasticity or litigants to plausibly characterize ordinary arbitral grievances as falling within the narrow window of review. As long as petitions can cross the threshold by invoking these open-textured standards, the supervisory shadow of the court inevitably re-enters the arbitral process. The current jurisprudence therefore reflects a structural mismatch: while it proclaims minimal intervention in theory, its vague threshold enables constitutional engagement in practice. A more effective reconciliation between arbitral autonomy and constitutional oversight will require the courts to articulate clearer, objectively administrable markers for when writ jurisdiction may truly be invoked mid-arbitration.
Indian Jurisprudence in Contrast with International Standards
Conducting a comparative analysis across jurisdictions reveals a two-step pattern that is important for the present argument. First, legal systems that are strongly pro-arbitration typically confine court involvement during the arbitral process to narrowly drawn exceptions. Second, even where courts retain a supportive role, the conditions for intervention are framed in objective terms rather than open-ended standards. For this reason, United Kingdom and France are adopted as comparative benchmarks here. Both are recognized among the most preferred arbitral seats in the world. Even more important, is the fact that France offers a strict abstention model, while the UK offers a supportive but tightly channeled model. Taken together, they demonstrate a common comparative principle: whether the system is abstentionist or supportive, judicial restraint is specifically defined in advance, not left to broad discretionary review.
France offers a model of strict non-intervention, governed by the “Negative Effect of Kompetenz-Kompetenz.” Under Article 1448 of the Code of Civil Procedure, if an arbitration agreement exists, a French court must decline jurisdiction. The only two exceptions to this rule are if the arbitration agreement is manifestly null and void or manifestly unenforceable. Even so, Courts do not conduct a full merits inquiry but only examine whether the agreement is invalid or unenforceable. There is a very high threshold which ensures that courts only screen for obvious defects. Unlike India, where a judge might examine if a tribunal’s order is “perverse” under Article 226, a French judge cannot examine the merits of the jurisdiction at the pre-award stage at all. They only have the power to look into whether the arbitration clause is prima facie null and void. The standard is objective and superficial, stripping the court of the discretionary power that creates vagueness in the Indian context.
The United Kingdom adopts a ‘supportive’ but strictly defined approach on the other hand. Under the Arbitration Act 1996, court intervention is not based on vague concepts of “justice” but on the objective “Unable to Act” test. Section 44(5) dictates that the court shall act only if the arbitral tribunal “has no power or is unable for the time being to act effectively.” This is a binary, factual test: does the arbitration have the legal power to grant the remedy? If yes, the court is statutorily barred from intervening. Unlike India, which allows a court to intervene based on the quality of the tribunal’s decision, the UK standard only looks at the capacity of the tribunal to act, eliminating the risk of judicial overreach.
From an even broader perspective, although the Act declares in its preamble that it is enacted taking into account the UNCITRAL Model Law, the decision in Union of India v. East Coast Boat Builders & Engineers Ltd., makes it clear that the Indian legislature intentionally did not adopt Article 16(3) of the Model Law in its full form. Article 16(3) permits immediate judicial intervention where the arbitral tribunal rules on its own jurisdiction. However, as the Delhi High Court emphasized, the Model Law cannot be imported into the Indian statute where the Parliament has deliberately legislated otherwise. The Court observed that Sections 16 & 37 of the Act reflect a conscious departure from Article 16(3). The omission of any right to challenge a tribunal’s positive jurisdictional ruling at an interlocutory stage is not a legislative oversight or lacuna. Rather, it is a deliberate structural choice aimed at ensuring minimal judicial interference during ongoing arbitration. The judgment expressly rejects the argument that Model Law provisions or corresponding provisions in foreign statutes like the UK Arbitration Act can be used to create a right of premature court intervention where the Indian Act provides none. Thus, while the legislature consciously abandoned Article 16(3) to avoid early court intervention, the judicial practice has, in effect, reintroduced premature scrutiny of jurisdictional issues.
Proposed Reforms to Reinforce Minimal Court intervention in Indian Arbitration
Despite the textual clarity of Section 5 of the Act, Indian Arbitration continues to be derailed by writ petitions. The failure lies not only in the volume of intervention but in the judiciary’s formulation of vague carve-outs. Reform must begin by structurally foreclosing this backdoor.
First, codification of a strict threshold for the exercise of writ jurisdiction during ongoing arbitral proceedings, making clear that pre-award intervention is to remain exceptional and available only in narrowly defined cases – where the arbitral process is non est, where there is demonstrable prima facie bad faith, or where there is a fundamental procedural defect causing irreparable prejudice that cannot be cured through the post-award framework. All of these measures do not require the Court to go into the merits of the dispute. Such a provision would not extinguish the constitutional remedy available under Articles 226/227 but would reinforce the arbitral policy of minimal interference and reduce tactical interlocutory challengers.
Second, arbitration rules and model clauses may validly strengthen party autonomy by expressly waiving rights of appeal or review in respect of procedural or administrative decisions within the arbitral process. A prime example is SIAC Rule 40.2, under which parties waive “any right of appeal or review… to any State court or other judicial authority” of Registrar decisions. However, such clauses should not be framed as excluding the High Court’s constitutional jurisdiction under Article 226, rather their utility lies in shaping judicial discretion to the extent that it does not need to go into the merits of the dispute by seeing that sophisticated parties have agreed to finality for procedural determinations. Courts should treat that as a weighty factor making it easier to make a distinction between a merit’s inquiry and a prima facie agreement inquiry.
Third, create dedicated benches of chambers for arbitration matters (as in France’s specialized international chamber or Singapore’s SICC) to foster consistency. A limited parallel may be drawn with the logic of the Commercial Courts Act, 2015, which reflects a legislative preference for specialized adjudicatory pathways in commercial matters. While the Commercial Courts regime is not uniformly regarded as successful in practice, the institutional logic behind the Act of specialist judges being “better equipped to dispose of the cases efficaciously,” can be applied here. On that logic, a specialized arbitration bench/roster may not by itself eliminate the interventionism, but it could significantly reduce forum-shopping and gradually crystallize a more disciplined jurisprudence aligned with the aim of the Act.
*Aditya Mudhana is a second-year law student at Jindal Global Law School.