亚洲争议评论：无限仲裁豁免权之反思-Dongguan Arbitration Commission

Research

亚洲争议评论：无限仲裁豁免权之反思

Release Date：2025-09-22

This article was originally published on Asian Dispute Review, issue 2, 2025.

This article is an edited version of a prizewinning entry in HKIAC’s HK45 Essay Competition 2024.*

Examining the justifications proffered for arbitral immunity, the author argues that (1) both unlimited arbitral immunity and unlimited arbitral liability are untenable for public policy reasons, and (2) only a qualified arbitral immunity model, whereby arbitrators may be held liable only for intentional wrongdoing and gross negligence, appropriately balances the objectives of preserving the integrity of the arbitration process, ensuring the finality of arbitral awards and vindicating wronged parties’ rights to justice.

Introduction

It is a fundamental cornerstone of any system of justice that where there is a wrong, there should be a remedy. [1] While most jurisdictions recognise an exception to this principle in conferring varying degrees of immunity on arbitrators for civil liability arising out of the performance of their arbitral functions, [2] there is no international consensus on the limits of arbitral immunity. [3]

At one end of the spectrum, some jurisdictions extend unlimited immunity to arbitrators on the basis that arbitrators are like judges, who enjoy judicial immunity. [4] At the other end, some jurisdictions treat arbitrators in the same way as other professionals, without any special immunity. Most common law jurisdictions have trodden the middle ground in adopting qualified arbitral immunity, whereby arbitrators generally enjoy immunity, subject to limited exceptions. [5]

Judges and arbitrators: a flawed analogy

Arbitral immunity has traditionally been justified by analogy to judicial immunity, owing to the perceived functional equivalence between judges and arbitrators. [6] Per Lord Salmon in Sutcliffe v Thackrah, “arbitrators are in much the same position as judges, in that they carry out more or less the same functions.” [7] A similar line of reasoning has also been adopted in other common law jurisdictions, such as Hong Kong [8] and in the US, [9] where the concept of judicial immunity is deeply entrenched.

“While most jurisdictions … confer[] varying degrees of immunity on arbitrators for civil liability arising out of the performance of their arbitral functions, there is no international consensus on the limits of arbitral immunity.”

However, arbitrators are not judges. First, judges are accountable to the State, which has appointed them to administer justice, not to the parties involved in the dispute before them. [10] Citizens do not have the power to select judges, nor do they pay them directly. [11] Crucially, judges do not owe litigants any legal duty of care. [12] By contrast, arbitrators are freely appointed by the parties, who pay them for their services. Thus, arbitrators are accountable to the parties (and arbitral institutions, where applicable) in the exercise of their functions and generally owe duties to act with due care, to act promptly and to act judicially. [13] Given this accountability and the fact that they are selected by the parties for their professed technical or intellectual expertise, arbitrators should not have unlimited immunity. [14]

“Arbitral immunity has traditionally been justified by analogy to judicial immunity, owing to the perceived functional equivalence between judges and arbitrators. … However, arbitrators are not judges.”

Second, arbitrating parties generally have no right to appeal against an arbitral award [15] and can only challenge it in limited circumstances. [16] Conversely, most (if not all) jurisdictions provide for the right to appeal against the decision of a judge. [17] Thus, challenging an award is more difficult than appealing a court judgment.

“… [A]rbitrators are accountable to the parties (and arbitral institutions, where applicable) in the exercise of their functions and generally owe duties to act with due care, to act promptly and to act judicially. Given this accountability and the fact that they are selected by the parties for their professed technical or intellectual expertise, arbitrators should not have unlimited immunity.”

Third, judges operate under public scrutiny, whereas arbitral proceedings are confidential. Judicial decisions have “far-reaching and precedential consequences”, [18] while arbitrators do not create precedents and awards are confidential to the parties involved. Thus, by comparison to judges, there are fewer incentives for arbitrators to adhere to the proper exercise of their arbitral functions.

Immunity should not, therefore, extend to arbitrators on the sole basis that they appear to have a similar role to judges. Independent policy reasons must exist to justify a derogation from the fundamental principle that a wronged party should have a remedy against the wrongdoer.

Policy justifications for arbitral immunity

(1) Independence and integrity of the arbitral process

One of the primary policy justifications for arbitral immunity is that it protects the independence and integrity of the arbitral process, as arbitral immunity ensures that an arbitrator can make decisions without the fear of reprisals or harassment from an unsuccessful party. [19] While it is of paramount importance that arbitrators are not influenced in their decisions by a potential finding of liability, it is submitted that the concern that an arbitrator’s independent decision-making would be affected by the lack of immunity is overstated.

“One of the primary policy justifications for arbitral immunity is that it protects the independence and integrity of the arbitral process, as arbitral immunity ensures that an arbitrator can make decisions without the fear of reprisals or harassment from an unsuccessful party.”

First, it is highly unlikely that arbitrators, as professionals, would breach their duty to act independently by preferring one party over another out of the fear of reprisals or harassment from the unsuccessful party. In particular, many arbitrators are part-time or former judges who would be expected to act with integrity even when faced with the potential of a backlash for their decisions. It is also unlikely that arbitrators would risk their reputations by acting in a biased manner, especially since this would affect their prospects of securing future appointments.

Second, the prospect of being sued for negligence is unlikely to tempt professionals, such as doctors or lawyers, who do not need any incentives to comply with the rules and ethics of their respective professions, to disregard their duties. There is no reason why an arbitrator should be expected to act with any less professionalism than other professionals.

“While it is of paramount importance that arbitrators are not influenced in their decisions by a potential finding of liability, it is submitted that the concern that an arbitrator’s independent decision-making would be affected by the lack of immunity is overstated.”

Third, as Lord Kilbrandon observed in Arenson, [20] the distinction between an arbitrator and a mutual valuer appointed to adjudicate an issue between parties with competing interests is not clear. Both valuers and arbitrators are persons selected by the parties for their technical or intellectual expertise, of which they pledge skill in the exercise thereof. It is therefore difficult to see why arbitrators should be immune from suit if valuers are not.

Fourth, the concern that an arbitrator’s decision-making process may be affected by the prospect of liability may be mitigated by professional indemnity insurance. [21] While this would not generally apply to deliberate wrongful acts, [22] there is no reason why the law should protect arbitrators from liability in these circumstances, as explained further below.

(2) Skilled persons would be deterred from acting as arbitrators

Another oft-cited justification for arbitral immunity is that the absence of immunity would deter skilled persons from acting as arbitrators. [23] This is unconvincing.

First, the possibility of liability does not deter other professionals, such as doctors or lawyers, from practising. Again, there is no reason why arbitrators should be treated any differently from other professionals. Moreover, given the rise of arbitration as a profession, it is “implausible” [24] that imposing liability on arbitrators would “cause people to panic and desert this profession in flocks.” [25]

“Another oft-cited justification for arbitral immunity is that the absence of immunity would deter skilled persons from acting as arbitrators. This is unconvincing.”

Second, the availability of professional indemnity insurance would alleviate an arbitrator’s concerns about being potentially liable for substantial damages. As arbitrators continue to develop as a growing profession, there is no reason why professional indemnity insurance should not be available to them.

“… [T]he justification that arbitral immunity ensures the finality of arbitral awards is well-founded. Finality is one of the primary objectives of arbitration[.] … [B]ring[ing] claims against arbitrators on a whim and ‘reopen[ing]’ arbitration proceedings … is … undesirable as a matter of public policy.”

(3) Ensuring finality of arbitral awards

By contrast, the justification that arbitral immunity ensures the finality of arbitral awards is well-founded. [26]

Finality is one of the primary objectives of arbitration, [27] and parties who have chosen arbitration over litigation would likely have done so because arbitration is said to be a more efficient process that facilitates the resolution of disputes in a speedier and less costly manner. However, if arbitrating parties were able to bring claims against arbitrators on a whim and ‘reopen’ arbitration proceedings, their ability to draw proceedings to a close and put their dispute behind them would be lost. This would likely have a detrimental effect on the development of international arbitration as an alternative to litigation and would therefore be undesirable as a matter of public policy.

“As arbitral immunity has a deterrent effect on potential litigants, it would reduce the number of meritless claims brought against arbitrators as a tool for undermining arbitral awards and would therefore ensure the finality of awards.”

Moreover, in the absence of arbitral immunity, unsuccessful parties may bring meritless claims against arbitrators in an attempt to attack the awards rendered and to avoid compliance with them. As commentators have observed, this appears to be a growing trend in recent years. [28] As arbitral immunity has a deterrent effect on potential litigants, it would reduce the number of meritless claims brought against arbitrators as a tool for undermining arbitral awards and would therefore ensure the finality of awards.

Policy justifications against unlimited arbitral immunity

While ensuring the finality of arbitral awards justifies arbitral immunity, it does not do so to an unlimited extent. Unlimited arbitral immunity would render arbitrators impervious and unaccountable for any misconduct, no matter how morally reprehensible. This would be problematic, for the following reasons.

“While ensuring the finality of arbitral awards justifies arbitral immunity, … [u]nlimited arbitral immunity would render arbitrators impervious and unaccountable for any misconduct, no matter how morally reprehensible. This would be problematic[.]”

First, if arbitrators can escape with impunity for any wrongdoing, [29] they may be encouraged to become careless and to be more likely to render poor decisions. [30] Although codes of ethics and guidelines exist to regulate arbitrators’ conduct and may, in theory, militate against this, their impact in holding arbitrators accountable is, in the absence of enforcement, questionable. [31] Moreover, there are generally no professional associations of arbitrators that discipline arbitrators for misconduct, so that an arbitrator’s abuse of his or her discretion may go unchecked. [32] This is detrimental not only to the arbitrating parties, but also to the integrity of the international arbitration system.

Second, to deny a wronged party a remedy against an arbitrator for their wrongdoing offends the fundamental principle that where there is a wrong, there should be a remedy. Per Lord Dyson in Jones v Kaney, [33] this should always be exceptional and any justification for doing so must be necessary. However, the justifications proffered for arbitral immunity do not demonstrate such a necessity. As explained above, the functional equivalence between judges and arbitrators is dubious at best, and most of the policy justifications for arbitral immunity are more apparent than real. In any event, concern that the finality of awards may be upset can be accommodated by qualified arbitral immunity, thus obviating the need for unlimited arbitral immunity, as explained further below.

The case for qualified arbitral immunity and its proper limits

Both unlimited immunity and unlimited liability are untenable options. The former would leave a wronged party without any recourse in all circumstances, while the latter would likely lead to a floodgate of claims against arbitrators and the unravelling of arbitral awards. By contrast, qualified immunity strikes an appropriate balance between the objectives of ensuring the finality of arbitral awards and vindicating a wronged party’s rights to justice.

“Both unlimited immunity and unlimited liability are untenable options. The former would leave a wronged party without any recourse in all circumstances, while the latter would likely lead to a floodgate of claims against arbitrators and the unravelling of arbitral awards.”

What, then, should the proper limits of qualified arbitral immunity be? Even among jurisdictions that have adopted qualified arbitral immunity, the limits of liability have been drawn differently. In England & Wales, an arbitrator is not liable for anything done or omitted to be done as arbitrator “unless the act or omission is shown to have been in bad faith.” [34] In Hong Kong, an arbitrator is liable “only if it is proved that the act was done or omitted to be done dishonestly.” [35] In Singapore, an arbitrator is only immune for (1) negligent acts done or omitted to be done in the capacity of an arbitrator, and (2) any mistake in law, fact or procedure made in the course of arbitral proceedings or in the making of an arbitral award, [36] thus leaving an arbitrator open only to liability for intentional wrongdoing and gross negligence. [37]

“… [Q]ualified immunity strikes an appropriate balance between the objectives of ensuring the finality of arbitral awards and vindicating a wronged party’s rights to justice. … [However,] [e]ven among jurisdictions that have adopted qualified arbitral immunity, the limits of liability have been drawn differently.”

Given the objective of ensuring the finality of arbitral awards, arbitrators should generally enjoy immunity, save in circumstances where there has been (1) intentional wrongdoing and/or (2) gross negligence on their part.

(1) Intentional wrongdoing

Arbitrators should be liable if they intentionally abuse their discretion or otherwise engage in bad faith conduct in carrying out their arbitral functions. Allowing arbitrators who intentionally act in bad faith to escape with impunity would amount to “legalised fraud”, [38] which damages the reputation and integrity of international arbitration. Parties may also be less inclined to arbitrate their disputes as no recourse would be available to them even where an arbitrator has deliberately acted in bad faith. Unlike under national court systems, there are no equivalent safeguards in place to ensure proper arbitral conduct.

“Given the objective of ensuring the finality of arbitral awards, arbitrators should generally enjoy immunity, save in circumstances where there has been (1) intentional wrongdoing and/or (2) gross negligence on their part.”

Further, a primary function of the law is to prevent undesirable behaviour and secure desirable behaviour. [39] As a matter of public policy, the law should not condone intentional wrongdoing. Thus, most jurisdictions do not allow a party to exclude liability for intentional wrongdoing, such as fraud. [40]

Finally, allowing claims to be brought against arbitrators for intentional wrongdoing is unlikely to have a substantive effect on the finality of arbitral awards. A high threshold must be met to prove intentional wrongdoing, and recalcitrant parties would be less likely to bring baseless claims against arbitrators where a higher degree of proof is required. This is borne out by the fact that, notwithstanding the possibility of suing an arbitrator for bad faith in England & Wales, there are no reported English cases on this issue, which suggests that unsuccessful parties are not rushing to bring claims of bad faith against arbitrators to court. [41]

(2) Gross negligence

Arbitrators should also be liable if they have been grossly negligent in carrying out their arbitral functions. Although there is no universal definition of ‘gross negligence’, it is generally accepted that gross negligence involves a “serious disregard of or indifference to an obvious risk.” [42]

Although intentional wrongdoing is arguably more morally reprehensible than gross negligence (in that the latter is concerned with competence and not an intention to commit a wrong), gross negligence is also morally reprehensible. [43] Thus, the common law recognises that gross negligence is sufficient to found criminal liability, whereas mere negligence generally is not. [44]

“Arbitrators should be liable if they intentionally abuse their discretion or otherwise engage in bad faith conduct in carrying out their arbitral functions. [However, a] high threshold must be met to prove [it] … and recalcitrant parties would be less likely to bring baseless claims against arbitrators where a higher degree of proof is required.”

Moreover, while it is arguable that negligence is also morally reprehensible, given that negligence is easier to prove than gross negligence, unsuccessful parties would be more likely to bring claims against arbitrators if the latter could be held liable for mere negligence. Thus, the concern that the finality of arbitral awards would be upset would be likely to materialise in such circumstances.

Conclusion

The right to a remedy is a fundamental right enshrined in the common law, dating back to the early 18th century. [45] To deprive a wronged party of a remedy against the wrongdoer therefore requires strong and cogent justification.

“Although intentional wrongdoing is arguably more morally reprehensible than gross negligence (in that the latter is concerned with competence and not an intention to commit a wrong), gross negligence is also morally reprehensible.”

The author has argued that no such justification exists for affording unlimited immunity to arbitrators. The traditional justification for arbitral immunity based on the functional equivalence between arbitrators and judges does not withstand scrutiny. Further, an examination of the policy arguments for arbitral immunity demonstrates that only ensuring the finality of arbitral awards justifies arbitral immunity.

“The right to a remedy is a fundamental right enshrined in the common law, dating back to the early 18th century. To deprive a wronged party of a remedy against the wrongdoer therefore requires strong and cogent justification.”

However, unlimited arbitral immunity is untenable, as a wronged party would have no recourse against the wrongdoer in any circumstances, even if the wrongdoer were guilty of morally reprehensible conduct which, as a matter of public policy, the law should not condone. Unlimited arbitral liability is equally untenable as this would likely lead to a floodgate of claims against arbitrators and hinder the finality of arbitral awards.

“Ultimately, the question of arbitral immunity is a matter of public policy, which may evolve over time. At present, there is no reason for extending unlimited immunity to arbitrators.”

The author therefore makes the case for qualified arbitral immunity, as this better balances the objectives of ensuring the finality of arbitral awards and vindicating a wronged party’s rights to justice. Arbitrators should generally be immune, except in cases of intentional wrongdoing and gross negligence, thus safeguarding the integrity of the arbitration process and the rights of arbitrating parties. Ultimately, the question of arbitral immunity is a matter of public policy, which may evolve over time. At present, there is no reason for extending unlimited immunity to arbitrators.

References (Please swipe up and down to see)

* As to the HK45 Essay Competition 2024, see https://hkiac.glueup’com/event/hk45-2024-essay-competition-103425/details.html. The topic selected by the author (the winner in the Global/Hong Kong Category) was Option 3, ‘Should an arbitrator have unlimited arbitral immunity?’ The winning entry on the same topic in the Emerging Economies Category, by Ananya Sahu, will be published in the July 2025 issue of Asian Dispute Review.

[1] Jones v Kaney [2011] UKSC 13, at [113], per Lord Dyson (UK Supreme Court). See also infra, p 88 and note 45.

[2] Whether arbitrators should be held criminally liable for their conduct in performing their arbitral functions is beyond the scope of this article.

[3] Susan D Franck, The Liability of International Arbitrators: A Comparative Analysis and Proposal for Qualified Immunity, 20 NYLS J Int’l & Comp L 1-60 (2000), p 31.

[4] Judicial immunity is a concept developed in the common law which protects judges from personal liability in respect of any actions performed in their judicial capacity, even if those actions are done maliciously or in bad faith. See Sirros v Moore & Ors [1975] QB 118, at 137D-F, per Lord Denning MR (Court of Appeal, England & Wales).

[5] Franck, op cit (note 3), pp 33-36.

[6] Lord Dyson, The Proper Limits of Arbitrators’ Immunity (2018) 84(3) Arbitration: The International Journal of Arbitration, Mediation and Dispute Management 196-204, at p 196.

[7] Sutcliffe v Thackrah [1974] AC 727 (House of Lords).

[8] In Song Lihua v Lee Chee Hong (formerly Que Wenbin) [2023] HKCFI 1954 at [31], it was held in the Hong Kong Court of First Instance that “arbitrators should be entitled to the same immunity available to judges in respect of their decision-making in the process of arbitration, absent fraud or bad faith.” (Per Mimmie Chan J)

[9] In Hudnall v Smith 2021 US Dist LEXIS 162482 at [4] (US District Court, WD Texas, 10 August 2021), the court stated that “[t]he resemblance of an arbitrator’s role to that of a judge gives rise to immunity from suit similar to judicial immunity.” (Per Judge Montalvo).

[10] Arenson v Casson Beckman Rutley & Co [1977] AC 405, at 431F-H, per Lord Kilbrandon (House of Lords).

[11] Ibid, at 431G.

[12] Ibid, at 432A.

[13] Redfern and Hunter on International Arbitration (7^thEdn, 2022) (Redfern and Hunter), paras 5.54-5.85.

[14] Supra (note 10), at 431B, per Lord Kilbrandon.

[15] Dyson, op cit (note 6), p 198.

[16] For example, in Hong Kong, although an award may be challenged under the ‘opt in’ provisions of ss 99 and Sch 2, ss 4-6 to the Arbitration Ordinance (Cap 609) on the basis of serious irregularity or on a question of law, a high threshold must be met for a challenge to succeed. Thus, even if an arbitrator has erred in law, an award cannot be set aside unless the error is egregious and causes substantial injustice: see LY v HW [2022] HKCFI 2267.

[17] Dyson, op cit (note 6), p 198.

[18] Baar v Tigerman, 189 Cal Rptr 834 (California Court of Appeals, 17 March 1983).

[19] Redfern and Hunter, supra (note 13), para 5.65. See also Sutcliffe v Thackrah, supra (note 7) at 736, in which Lord Reid explained that “an arbitrator might be influenced by the thought that he was more likely to be sued if his decision went one way than if it went the other way”, and Song Lihua v Lee Chee Hong (formerly Que Wenbin), supra (note 8), in which Mimmie Chan J stated (at [35]) that “[s]uch immunity is an essential foundation for judicial and arbitral integrity and independence, to ensure that arbitrators and judges can make their decisions on the right result without fear or distractions as to whether they could be made liable for claims of any party.”

[20] Supra (note 10), at 432-433.

[21] As the ICC Commission on International Arbitration’s Working Party on the status of the arbitrator recommended, in its Final Report on the Status of the Arbitrator (1996) 7(1) ICC International Court of Arbitration Bulletin at p 31, it was “appropriate for civil liability insurance to be taken out by the arbitrator or, in the case of institutional arbitration, by the arbitration centre on behalf of the arbitrator.”

[22] Ibid.

[23] Redfern and Hunter, supra (note 13), para 5.66.

[24] Jason Yat-Sen Li, Arbitral Immunity: A profession comes of age (1998) 64(1) Arbitration: The International Journal of Arbitration, Mediation and Dispute Management 51-57, p 55.

[25] Ibid.

[26] Redfern and Hunter, supra (note 13), para 5.66.

[27] Franck, op cit (note 3), p 54.

[28] Paula Hodges, Time to Stand Up for Immunity (2023) 39(2) Arb Int’l 234-245, p 234.

[29] In the US, arbitral misconduct merely prevents arbitrators from collecting their fees. See Franck, supra (note 3), p 32.

[30] Redfern and Hunter, supra (note 13), para 5.67.

[31] Maureen A Weston, Reexamining Arbitral Immunity in an Age of Mandatory and Professional Arbitration, 88 Minn L Rev 449-517 (2004), p 458.

[32] Franck, op cit (note 3), p 54.

[33] Jones v Kaney, supra (note 1), at [108].

[34] Section 29(1) of the English Arbitration Act 1996.

[35] Section 104 of the Hong Kong Arbitration Ordinance (Cap 609).

[36] Section 25 of the Singapore International Arbitration Act (Cap 143A, 2002 Rev Ed) and s 20 of the Singapore Arbitration Act (Cap 10, 2002 Rev Ed).

[37] Franck, op cit (note 3), p 36.

[38] Ibid, p 58.

[39] Joseph Raz, ‘The Functions of Law’, in The Authority of Law: Essays on Law and Morality (2nd Edn, 2009, Oxford University Press and 2012, Oxford Academic online), ch 9, p 169.

[40] For example, under English law, a contracting party cannot exclude its liability for fraud. See HIH Casualty and General Insurance Ltd & Ors v Chase Manhattan Bank & Ors [2003] UKHL 6, at [16], per Lord Bingham of Cornhill (House of Lords): “It is clear that the law, on public policy grounds, does not permit a contracting party to exclude liability for his own fraud in inducing the making of the contract.”

[41] Hodges, op cit (note 28), p 240.

[42] Red Sea Tankers Ltd & Ors v Papachristidis & Ors, The Hellespont Ardent [1997] 2 Lloyd’s Rep 547, at 586, per Mance J (Commercial Court, England & Wales).

[43] Culpable negligence should also be classified as a serious moral wrong; this is, however, beyond the scope of this article. See Seana Valentine Shiffrin, ‘The Moral Neglect of Negligence’, in David Sobel, Peter Vallentyne & Steven Wall (Eds), Oxford Studies in Political Philosophy, Vol 3, ch 8 (2017, Oxford University Press and Oxford Academic online).

[44] Gross negligence, for example, is a criminal offence at common law: Adomako v R [1994] UKHL 63 (House of Lords). See note 2, supra.

[45] In Ashby v White (1703) 92 ER 126 at 134 (Court of Queen’s Bench), Sir John Holt CJ (dissenting) stated that “[i]f the plaintiff has a right, he must of necessity have a means to vindicate and maintain it, and a remedy if he is injured in the exercise or enjoyment of it, and, indeed it is a vain thing to imagine a right without a remedy; for want of right and want of remedy are reciprocal …” Editorial note: On appeal to the House of Lords ((1704) 17 HLJ 369), the judgment of the lower court was reversed on the basis of the Chief Justice’s dissenting judgment.

注：以上文章来源Asian Dispute Review，不代表东仲立场或观点，如您对转载本文有任何疑问，请及时与我委联系（电话：0769-28820008)