“Reasonable Possibility” or “Real Danger”: Exploring the Standards Used to Determine Arbitral Bias

Introduction

Independence and impartiality are the touchstones of an ideal arbitral proceeding. However, arbitrators are often selected from a limited pool of experts in a particular field, and parties may intentionally appoint one who has delivered favourable awards in the past. Hence, the possibility of bias towards the appointing party cannot be ruled out, thus undermining the legitimacy of the arbitral process. This issue is exacerbated in ad hoc arbitrations conducted without institutional oversight.

Bias may be of three kinds: apparent i.e. undue partiality towards one party; systemic, based on laws and institutions designed to inherently favour certain outcomes; and ideological, based on one’s social, economic, and political views.¹ However, the jurisprudence regarding arbitral bias is primarily focused on apparent bias, since it is apparent bias that affects the outcome of the dispute most directly. Hence, for the purposes of this article, we shall be using “bias” to refer to apparent bias.

Despite some consensus on the meaning of bias, countries use different tests and thresholds to determine the presence of bias. While this approach works for domestic arbitration, it can pose challenges for international arbitration due to lack of uniform standards. This difficulty can be highlighted using the decision in Avitel Post Studioz Ltd. v. HSBC PI Holdings (Mauritius) Ltd.² (Avitel), where the Supreme Court of India (the Court) directed courts to adopt international best practices instead of domestic standards while determining bias in cases arising out of international commercial arbitration.

The Avitel case

The present dispute arose out of a share subscription agreement (SSA) entered between HSBC PI Holdings, Mauritius (HSBC) and Avitel Post Studioz Limited (Avitel India). Pursuant to the SSA, HSBC invested USD 60 million in Avitel India to acquire 7.8% of its equity capital. This SSA contained an arbitration clause which provided that the disputes shall be resolved at the Singapore International Arbitration Centre (SIAC).

HSBC claimed that Avitel India had represented that this investment would be required to service a significant contract with the British Broadcasting Corporation (BBC contract). However, HSBC subsequently discovered that the BBC contract was non-existent, and the invested amount had been siphoned off to different companies by Avitel India. Hence, HSBC invoked arbitration, alleging fraudulent misrepresentation. Following this, an arbitral award (award) was passed, directing Avitel India to pay USD 60 million as damages to HSBC. After years of other litigation on multiple issues, including a landmark decision on the arbitrability of fraud, the present appeal was filed before the Court. Avitel India had challenged the enforceability of the award, on the grounds that the arbitrator was biased and hence, enforcement of the award would be contrary to the public policy of India under Section 48(2)(b) of the Arbitration and Conciliation Act, 1996 (A&C Act).

Avitel India argued that the presiding arbitrator, Mr Lau had failed to make a full and frank disclosure of material facts and circumstances concerning conflict of interest, referring to his holding some shares in a company that had contractual relations with HSBC Singapore. They submitted that General Standard 3 of the IBA Guidelines on Conflicts of Interest in International Arbitration, 2004 (IBA Guidelines) was violated since the independence and impartiality of the presiding arbitrator was compromised. Avitel India also attempted to raise an additional challenge to the award under Section 48(1)(b) of the A&C Act on account of “inability to present their case”. In response, HSBC contended that HSBC Mauritius, being the award-holder and the HSBC Singapore were different companies, and further, the relationship of an unrelated company with HSBC Singapore would not constitute bias with regards to HSBC Mauritius. Additionally, HSBC submitted that Mr Lau was not an employee of the other company; therefore, he was not incapacitated in his judgment as an arbitrator. HSBC also raised a procedural defence that the issue of bias could not be raised for the first time at the stage of enforcement of the award.

The Court observed that India was one of the earliest signatories to the Convention on the Recognition and Enforcement of Foreign Arbitral Awards, 1958 (New York Convention), which had superseded the Geneva Convention of 1927. Article V(2)(b) of the New York Convention provides that the recognition and enforcement of an arbitral award may be refused if the competent authority in the country where recognition and enforcement is sought finds that such an act would be contrary to the public policy of that country. The United States Court of Appeals, Second Circuit had held in Parsons & Whittemore Overseas Co. Inc. v. Societe Generale de L’Industrie du Papier³ that the New York Convention’s public policy defence should be construed narrowly, and enforcement of foreign arbitral awards may be denied only where enforcement would violate the forum State’s most basic notions of morality and justice. This understanding was incorporated in Indian jurisprudence through Renusagar Power Co. Ltd. v. General Electric Co.⁴ (Renusagar) and the A&C Act.

The Court noted that other countries such as France followed international dimensions instead of their domestic ones when applying public policy to New York Convention awards. However, Article V(2)(b) of the New York Convention necessitated the evaluation of the award as per the public policy of the country where the award was enforced, not the country where the award was passed. The Court in Renusagar case⁵ had noted that there was no uniform understanding of what constituted international public policy, and hence it sought to lay out a definition for “public policy of India” basis which to judge the arbitral awards. Eventually, the Court in Shri Lal Mahal Ltd. v. Progetto Grano SpA⁶ had held that the wider interpretation given to “public policy of India” in the domestic sphere under Section 34(2)(b)(ii) of the A&C Act would not apply to the enforcement of the international awards under Section 48(2)(b) of the A&C Act. Hence, the scope of “public policy” as a ground for resisting enforcement of a foreign award is narrower than the scope for challenging a domestic award. The Court also acknowledged that this view was favoured by the International Law Association as far back as 2002. Thus, in Avitel case⁷, the Court concluded that since India was a signatory to the New York Convention, it had to adopt an “internationalist approach” by clearly distinguishing between the standards applicable for domestic arbitration and international commercial arbitration for public policy, including bias.

The Court was cautious in this approach and highlighted that there was no single international test to decide allegations of bias and that different thresholds existed across different jurisdictions, like “real possibility of bias” in the UK, “real danger of bias” in Australia, and “reasonable suspicion” in Singapore. However, the Court also noted that embracing international standards would help build trust and provide certainty in dispute resolution. Finally, the Court held that the most basic notions of morality and justice under the concept of “public policy” would include bias, which had to be adjudicated based on stringent international best practices instead of wider domestic standards. It is only in exceptional circumstances that enforcement of an award may be refused on the ground of bias.

The Court also considered the IBA Guidelines, which though not binding, are relied upon by courts globally to determine bias, and were largely incorporated into the A&C Act in the Fifth and Seventh Schedules respectively. The IBA Guidelines follow a system of classifying disclosure scenarios into red, orange, and green lists to ensure consistency, and require an arbitrator to refuse appointment in case of any doubts as to impartiality or independence and disclose potential conflicts to parties. The High Court in Avitel case⁸ had applied the “reasonable third person” test contained in the IBA Guidelines to conclude that there was no requirement of disclosure by Mr Lau, therefore, the award could not be challenged on the grounds of arbitral bias.

The Court also noted that the parties had chosen Singapore as the seat of arbitration, yet the award was not challenged in Singapore when it was passed, but only in India at the time of its enforcement. It relied on multiple precedents to hold that the objection of bias must first be raised in the country of origin of the award, and the failure of Avitel India to do so in a timely fashion indicated that it was attempting to strategically delay the enforcement process. Hence, the appeal failed as the alleged bias did violate the most basic notions of morality and justice or shock the conscience of the Court. The Court also lamented HSBC’s arduous struggle to gather the fruits of the award, as enforcement had been delayed for over ten years. Hence, it categorically held that enforcement of foreign awards should only be refused in rare cases where non-adherence to international standards was clearly demonstrable.

Domestic thresholds

The Avitel case⁹ hinged on the difference between the narrow and strict international standards for determining bias, and the wide thresholds under domestic Indian law. This raises the question, what does this gap entail?

The A&C Act does not mention the term “bias”. However, Section 12(1)(a) of the A&C Act mandates a potential arbitrator to make disclosures regarding the existence of any relationship with the parties that could give rise to justifiable doubts. Section 12(1) provides as follows:

12. (1). When a person is approached in connection with his possible appointment as an arbitrator, he shall disclose in writing any circumstances,—

(a) such as the existence either direct or indirect, of any past or present relationship with or interest in any of the parties or in relation to the subject-matter in dispute, whether financial, business, professional or other kind, which is likely to give rise to justifiable doubts as to his independence or impartiality;

Section 12(3)(a) of the A&C Act further provides for a challenge to the arbitrator if he is not independent or impartial, as per the criteria laid down in Schedule V of the A&C Act, which provides for circumstances that give rise to justifiable doubts.

The Court in Voestalpine Schienen GmbH v. DMRC Ltd.¹⁰ had observed that the test was that of “apprehension of bias”, where the presence of any of the circumstances mentioned in the Fifth Schedule would give rise to a justifiable apprehension of bias. If this condition regarding apprehension of bias is met, Section 13(3) of the A&C Act provides for the arbitrator to recuse voluntarily, or for the Tribunal to decide on the challenge. If the arbitrator is not removed, the aggrieved party may challenge the award under Section 34 of the A&C Act.

Further, Section 12(5) of the A&C Act provides that a person whose relationship to the parties or their counsel falls under the list mentioned in Schedule VII of the A&C Act shall be ineligible to be an arbitrator in that case. The Court, in HRD Corpn. (Marcus Oil and Chemical Division) v. GAIL¹¹, also allowed parties to approach courts as an interlocutory mechanism, to decide on the ineligibility of an arbitrator under Schedule VII of the A&C Act.

The test developed in domestic jurisprudence to determine whether the arbitrator was biased or lacked impartiality is different from other standards. Though courts and tribunals follow the test of whether there is a “reasonable apprehension of bias”, the test in India is from the perspective of the party concerned rather than that of the arbitrator or a neutral third person. This was propounded by the Supreme Court in Ranjit Thakur v. Union of India¹², where it was held:

17. As to the tests of the likelihood of bias what is relevant is the reasonableness of the apprehension in that regard in the mind of the party. The proper approach for the Judge is not to look at his own mind and ask himself, however, honestly, “Am I biased?”; but to look at the mind of the party before him.

Though Ranjit Thakur case¹³ was regarding judicial bias, the same test has also been used to determine arbitral bias. However, it is unclear whether this standard is applied consistently since courts often refer to international standards while deciding on the issue of bias.

Is there an international standard?¹⁴

Though the Court in Avitel case¹⁵ declared that international standards must be followed for the purpose of determining arbitral bias, it is not certain what these standards are. Multiple institutions apply different tests, for instance, the United Nations Commission on International Trade Law (UNCITRAL), London Court of International Arbitration (LCIA) and SIAC use the test of whether there are “justifiable doubts” about the arbitrator’s independence or impartiality, while International Centre for Settlement of Investment Disputes (ICSID) requires a “manifest lack of qualities” to be proved for the arbitrator concerned. However, apart from arbitral institutions and conventions, countries too have adopted different standards, as was recognised by the Court in Avitel case¹⁶.

How then, does one determine which tests must be applied by Indian courts while deciding on bias during enforcement of international commercial awards? One view might be to adopt the test most used by common law jurisdictions. On the other hand, it could be argued that more recent and well-developed thresholds should be adopted since Indian courts have the unique opportunity to carve a fresh path. It can also be contended that the IBA Guidelines should provide the template for subsequent Indian case law development since they were relied upon and implicitly affirmed in Avitel case¹⁷. To that end, it is important to note that while Avitel case¹⁸ had relied upon the 2004 IBA Guidelines, they have been revised in 2024 based on the decision in Halliburton Co. v. Chubb Bermuda Insurance Ltd.¹⁹ (Halliburton).

Halliburton case²⁰ was decided by the Supreme Court of the United Kingdom in 2020 on an application for removal of an arbitrator for being biased. The judgment affirmed that the test was that of apparent bias i.e. whether a fair-minded and informed observer would conclude that there is a “real possibility” of bias, or the appearance of bias, even though the arbitrator might not have been actually biased. The Supreme Court of the United Kingdom noted that an arbitrator has a duty to disclose facts and circumstances which would or might reasonably give rise to the appearance of bias, to maintain transparency in arbitration and align with the best practices set out in the IBA Guidelines.

This was affirmed in the recent case of H1 v. W²¹ (H1) where the High Court of Justice of England and Wales allowed an application for removal of an arbitrator. In H1 case²², the arbitrator had made comments during proceedings where he indicated that one party’s experts were “exceptional people in their fields” and that he “knew them all personally extremely well”. The other party had thus asked for the arbitrator to be set aside on the grounds of apparent bias. The Court in this case observed that a fair-minded and informed observer would understand that experienced practitioner in a specific industry might know other experts who might be called in as witnesses, however, the remarks made by the arbitrator would lead this fair-minded and informed observer to the conclusion that there was a real possibility of him being biased. Hence, since it was held that an informed observer would conclude that there was an appearance of bias which met the threshold of “reasonable possibility” of bias, the High Court of Justice of England and Wales removed the arbitrator in H1 case²³.

A similar standard is followed in Singapore. The Singapore International Commercial Court in CFJ v. CFL²⁴ reaffirmed the objective test laid down in Halliburton case²⁵ of whether a fair-minded observer would have “reasonable suspicion” of bias. This test of a “reasonable apprehension” of bias is also followed in Ontario, Canada as held in Dufferin v. Morrison Hershfield²⁶.

However, Canada does not have a uniform law, as a result of which British Columbia has a different threshold, of whether there is “real danger” of bias, which applies to international commercial arbitrations seated there as well. This stringent standard for international commercial arbitration is followed in Australia as well, as was held by the Federal Court of Australia held in Sino Dragon Trading Ltd. v. Noble Resources International Pte. Ltd.²⁷

The test of “real danger” of bias as opposed to just “reasonable apprehension” makes it harder for parties to challenge arbitrators since it increases the burden of proof. It also relies on different perspectives: the former relies only on the perspective of the “observer” to determine real danger of bias, while the latter relies on the viewpoint of a “fair-minded and informed observer” to ascertain only the appearance of bias. India currently follows another test altogether, of whether there is a reasonable apprehension of bias from the viewpoint of the party concerned. These tests are fundamentally different and cannot be reconciled. Yet, Avitel²⁸ decision makes it clear that an international standard need to be followed to limit challenges and facilitate India’s pro-arbitration approach.

In this regard, the choice is between the tests of “reasonable possibility” of bias as followed by the UK and Singapore, or the “real danger” of bias test as followed in Australia. Since the Supreme Court of India evaluated the claims made in Avitel case²⁹ based on the disclosure requirements in the IBA Guidelines, it is reasonable to believe that the Court subscribes to the “real possibility” standard laid down in the IBA Guidelines and followed by the UK and Singapore. Concurrently, the Court had also stressed on the importance of higher thresholds in cases of arbitral bias to reduce the incidence of challenges, which might indicate a leaning towards the “real danger” standard. Hence, it is still unclear which test must be applied by Indian courts for cases of bias in international commercial arbitration. However, it is possible that the Court has intentionally left the door open to allow considerations of seat, governing law, and institutional rules of specific jurisdictions while deciding cases of multi-jurisdiction arbitration.

Conclusion

The decision in Avitel case³⁰ is the latest to demonstrate the Court’s non-interventionist and pro-enforcement stance. It is crucial because it emphasised the importance of arbitrator impartiality by classifying arbitral bias as an aspect of public policy. Further, Avitel case³¹ defined the contours of judicial oversight for determining bias by categorically providing for the adoption of global standards in cases of international commercial arbitration.

From a commercial perspective, this decision serves to reduce barriers from the context of ease of doing business in India, by evaluating arbitral bias along the lines of the “reasonable suspicion of bias” test, as also applied in the UK and Singapore, and recommended by the IBA Guidelines. Though the Court did not directly reject the higher threshold provided in the “real danger” test, it is likely that the Indian judiciary is seeking to align itself with the more arbitration friendly jurisdictions.