In the recent decision of Aroma Franchise Company Inc. et al v Aroma Espresso Bar Canada Inc. et al, 2023 ONSC 1827 (“Aroma”), the Ontario Superior Court of Justice provided important guidance in relation to the reasonable apprehension of bias for arbitrators in circumstances where an arbitrator is retained on multiple occasions by the same party or firm. While the court’s guidance, discussed below, should be carefully considered by lawyers and litigants, it also raises a number of further questions.


Aroma Espresso Bar Canada Inc. (“Aroma Canada”) was the master Canadian franchisee of, and therefore acted as middleman in relation to individual Aroma franchisees for, Aroma Franchise Company Inc. (“Aroma Franchisor”). A dispute arose between the two parties regarding their master franchise agreement, which resulted in an arbitration run by a sole arbitrator (the “First Arbitration”) under the International Commercial Arbitration Act, 2017 (given that the Aroma Franchisor was not a Canadian entity) seated in Ontario. Aroma Canada was for the most part the successful party.

However, while the First Arbitration was still in progress, and in fact significantly advanced, the arbitrator was retained by counsel for Aroma Canada as the sole arbitrator on another dispute (the “Second Arbitration”). Neither Aroma Canada nor the Aroma Franchisor was a party to the Second Arbitration.

Just prior to releasing his final award for the First Arbitration, the arbitrator emailed counsel for both parties. In his email, the arbitrator inadvertently copied a lawyer from the same firm as counsel for the respondent, although the inadvertently copied lawyer was not involved in the First Arbitration. This inadvertent inclusion raised a concern in the mind of counsel for Aroma Franchisor.

Through subsequent correspondence, in which the arbitrator acknowledged having inadvertently copied the incorrect counsel, the arbitrator disclosed that he had been retained as arbitrator in respect of the Second Arbitration some time into the First Arbitration. The arbitrator further noted that counsel for Aroma Canada in the First Arbitration had “involvement from time to time” in the Second Arbitration, rather than day-to-day carriage. The arbitrator also expressed the view that there was no overlap in the issues presented by the First Arbitration and the Second Arbitration, and that he was unaware of any connection between the parties in the two arbitrations.

Aroma Franchisor thereafter applied to set aside the arbitrator’s final award, as well as his costs awards, on the basis of a reasonable apprehension of bias stemming from his engagement in and non-disclosure of the Second Arbitration.

The Court’s Decision

In reviewing the set-aside application, the Court canvassed a number of key issues in arriving at its ultimate conclusion that the awards should be set aside and that a new arbitration should be conducted by a new arbitrator.

Disclosure of the Second Arbitration

First, the Court considered whether it was incumbent upon the arbitrator to disclose the Second Arbitration to Aroma Franchisor. Relying on Article 12 of the Model Law (as incorporated into the International Commercial Arbitration Act, 2017) as well as the IBA Guidelines on Conflicts of Interest in International Arbitration, the Court concluded that those authorities necessitated a careful consideration of the circumstances in order to determine whether disclosure was required (in other words, the answer was not immediately obvious based on a review of those authorities). To that end, the Court considered a number of factors, including the following:

  • The expectations of the parties in the selection of the arbitrator. A review of the parties’ contemporaneous correspondence at the time of the arbitrator’s selection revealed that the parties expected that, if a proposed arbitrator had previously been retained or engaged by either party, then that retainer or engagement needed to be disclosed at that time;
  • The extent to which there were any overlapping issues as between the two arbitrations. On this point, the Court observed that there were some overlapping issues (i.e. similar causes of action), which based on the United Kingdom Supreme Court’s (UKSC) decision in Halliburton Company v. Chubb Bermuda Insurance Ltd. [2020] UKSC 48, might give rise to an appearance of bias. However, in this case, there were limited overlapping issues, and the Second Arbitration did not involve a franchise dispute and was in a different industry, such that there was no appearance of bias; and
  • The fact that the arbitrator was a sole arbitrator in both arbitrations (and therefore controlled the outcome in both). The Court did not explore this issue in detail, although the balance of the Court’s analysis suggests that the obligation to disclose was heightened by the fact that the arbitrator exerted greater control over the outcome than he might have done in the context of an arbitral tribunal.

The Court then turned to a review of the applicable institutional rules, including the UNCITRAL Arbitration Rules and the ADRIC Code of Ethics, highlighting that those rules variously require disclosure in circumstances that “could reasonably give rise to justifiable doubts” (emphasis added) as to their impartiality or independence, as well as “might create an appearance of partiality or bias” (emphasis added). Again, although not stated explicitly, the Court’s analysis appears to implicitly suggest that the bar for disclosure is lower than the balance of probabilities.

Finally, the Court turned to Halliburton v Chubb (noted above), considered to be the United Kingdom’s leading case on arbitrator bias. Although not identical to this case, Halliburton involved a very similar scenario in which an arbitrator accepted appointments from the same party in multiple, overlapping cases, arising out of the same incident, without disclosure. While the arbitrator disclosed his prior appointments at the time he was retained in the arbitration at issue, he then did not disclose the subsequent appointment. Although the UKSC determined that the arbitrator should have disclosed the subsequent appointments, it went on to find that this did not create a reasonable apprehension of bias.

Based on the foregoing, the Court in Aroma determined that the arbitrator ought to have disclosed his appointment in the Second Arbitration to the parties in the First Arbitration.

Reasonable Apprehension of Bias

Turning to the key issue of whether there was a reasonable apprehension of bias, the Court observed that the test for determining whether there is reasonable bias in respect of a judge applies with equal force to an arbitrator, despite the fact that their functions differ in several respects: “[W]hat would an informed person, viewing the matter realistically and practically – and having thought the matter through – conclude. Would he think that it is more likely than not that [the decision-maker], whether consciously or unconsciously, would not decide fairly.” The Court also considered a number of significant factors, including the following:

  • The threshold for a finding of real or perceived bias is a high one, since it calls into question both the personal integrity of the adjudicator and the integrity of the administration of justice. The grounds must be substantial, and the onus is on the party seeking to disqualify to bring forward evidence to satisfy the test.
  • The presumption of impartiality is high.
  • The inquiry is objective and requires a realistic and practical review of all the circumstances from the perspective of a reasonable person. The courts will not entertain the subjective views of the parties in making such a determination.
  • A challenge based on reasonable apprehension of bias will not be successful unless there is evidence to support the allegation beyond a mere suspicion that the hearing officer would not bring an impartial mind to bear. Mere suspicion is insufficient.
  • When considering bias, context matters. Any review of an arbitrator’s conduct must be considered in context and not through the review of selected excerpts or specifically chosen terms, phrases, or questions posed.

As a result, the Court concluded that the determination of whether a reasonable apprehension of bias exists is extremely fact specific.

It is apparent from the Court’s decision that the high bar noted in the case law and relevant guidelines must be met in order to support a finding of a reasonable apprehension of bias; however, in the circumstances that bar was found to have in fact been met here. In particular, the Court highlighted a number of factors it considered relevant in reaching that conclusion, including the following:

  • In respect of the Second Arbitration, Aroma Canada had not tendered evidence in respect of a number of salient points, including how much the arbitrator was being paid, who had suggested the arbitrator’s appointment, who had reached out to the arbitrator to retain him, and whether the parties to one arbitration were aware of the other arbitration;
  • The optics of Aroma Canada’s lead counsel retaining the arbitrator in the Second Arbitration after the First Arbitration was underway. Although not explicitly stated, the Court’s observation on this point raises questions as to whether the courts are or will be concerned with counsel retaining an arbitrator on multiple occasions and any related objectives of doing so. In that regard, Aroma Franchisor argued that the mere proffering of money to the arbitrator via the Second Arbitration was itself fatal to the arbitrator’s impartiality, which argument the Court did not expressly reject;
  • The fact that the arbitrator was selected for the Second Arbitration despite Aroma Canada’s counsel not having any prior experience with him as an arbitrator prior to the First Arbitration, and despite the availability of other competent commercial arbitrators in Toronto; and
  • The parties’ pre-appointment correspondence (discussed above), in which both parties emphasized the importance of selecting an arbitrator without a pre-existing relationship with either party or their counsel.

Based on the foregoing considerations, the Court determined that there was a reasonable apprehension of bias. In the circumstances, Article 18 of the Model Law was found to have been violated, which qualified as grounds for set-aside pursuant to Article 34(2) of the Model Law. As a result, the Court set aside the awards in the First Arbitration and directed that a new arbitration be conducted by a new arbitrator.

Review of the Decision and Questions Raised

Given the significant impact of the Court’s decision to remit the matter back for an entirely new arbitration, Aroma raises several important issues for further consideration.

First, prior to Aroma, the Halliburton decision was (and still is) considered a persuasive authority in the international arbitration community, and was considered instructive to arbitration practitioners in Canada notwithstanding that it was not a binding authority.

In that regard, it bears noting that in Halliburton, the arbitrator had arguably engaged in conduct that would give rise to an even greater apprehension of bias – there, the arbitrator had accepted appointments in multiple, overlapping matters from the same party, all arising out of the same incident (the Deepwater Horizon incident). Nevertheless, the UKSC found that an objective observer would not have concluded that the arbitrator was biased.

In this case, Aroma Franchisor argued – and the Court appears to have accepted – that Halliburton was distinguishable on the basis that (1) the applicable UK legislation set a higher threshold for removing an arbitrator or setting aside an award – namely, the applicant must show that a substantial injustice has been or will be caused – and (2) the UK legislation did not contain a statutory duty of disclosure, unlike the Model Law.

That being said, this position appears to somewhat understate the relevance of the UKSC’s findings in Halliburton insofar as (1) the test applied by the UKSC for bias was effectively the same as that applied in Aroma, yet the UKSC reached the opposite conclusion (i.e., that there was no bias), and (2) the UKSC found a common law duty of disclosure, such that it was functionally equivalent to the Model Law’s statutory duty (as expressed in the Ontario legislation). Furthermore, in respect of the “substantial injustice” requirement set out in Halliburton, it bears noting that the Court in Aroma similarly observed a finding of real or perceived bias requires “substantial” grounds. In that regard, these thresholds are more similar than they might first appear.

As a result, Halliburton potentially could have been considered as a more persuasive authority in Aroma against a finding of a reasonable apprehension of bias and may be considered as such in future cases.

Second, Aroma’s emphasis on the parties’ expectations as articulated in their pre-appointment correspondence is, in some sense, unfair to the arbitrator, insofar as the Court’s analysis does not suggest that the arbitrator had any knowledge of that correspondence or the importance that the parties had placed on their chosen arbitrator having no business relationship with either party or their counsel.

This issue in fact appears to have been the most important to the Court’s ultimate determination. In that regard, there exists a tension between the Court’s emphasis on the parties’ expectations, and the pre-existing case law establishing that courts will not entertain the subjective views of the parties in assessing a claim of bias. In any event, whereas knowledge by the arbitrator of the parties’ expectations may have militated in favour of a finding of an apprehension of bias (i.e. knowing of the parties’ wishes but acting against them), the opposite is equally true that lack of such knowledge could arguably militate against such a finding.

Third, the Court’s comments regarding the selection of the arbitrator for the Second Arbitration raise an interesting question as to the frequency with which an arbitrator may be appointed by the same counsel or parties. This question is particularly important in specialized practice areas, such as construction law, where there are a limited number of arbitrators with the subject matter expertise and experience to adjudicate such disputes.

To reiterate, the Court was somewhat critical of the fact that the arbitrator was selected for the Second Arbitration despite Aroma Canada’s counsel not having any prior experience with him as an arbitrator, and despite the noted availability of other commercial arbitrators in Toronto.

One the one hand, and as the Court observed, the IBA Guidelines identify three or more appointments by the same counsel within a period of three years as falling within the “orange list” as a problematic-but-not-disqualifying circumstance which may warrant disclosure; in other words, repeated use of an arbitrator may pose problems with respect to using them again. On the other hand, however, the Court appeared to be critical of the fact that Aroma Canada’s counsel had appointed the arbitrator a second time despite having had no experience with him as an arbitrator prior to the First Arbitration.

In the circumstances, these two propositions appear to be in tension insofar as they suggest a fairly narrow range of permissibility – it is potentially problematic to appoint an arbitrator whom counsel has already retained repeatedly, yet it may also be problematic to repeatedly appoint an arbitrator whom counsel has not already retained previously.

Fourth, the Court’s observations as to the optics of Aroma Canada’s lead counsel retaining the arbitrator in the Second Arbitration after the First Arbitration was underway – what the Court referred to as a “bad look” – raises an interesting question as to the presumption of an arbitrator’s impartiality. As noted above, the Aroma Franchisor appears to have argued that the fact money was proffered to the arbitrator via the Second Arbitration was in itself fatal to his role in the First Arbitration, while the balance of the judgment suggests a concern regarding the optics of counsel’s intentions and objectives in selecting the same arbitrator twice.

This raises questions for future decisions as to how the presumption of the arbitrator’s impartiality will be considered, as the Court did not explicitly confirm that the proffering of money is insufficient to ground a finding of bias. To the contrary, it can and should be presumed – which presumption can be rebutted with specific evidence – that the arbitrator will continue to act impartially in such circumstances; arbitration invariably involves remunerating arbitrators, and as such, the presence of remuneration should not in and of itself be disqualifying. Put differently, payment for services rendered by an arbitrator should not be considered the functional equivalent of an inducement.

Practically speaking, in specialized industries, it is not uncommon for a party to appoint the same arbitrator on matters while that arbitrator is already arbitrating prior matters in which the same party is involved. If the use of arbitrators on multiple construction matters were to qualify as grounds for reasonable apprehension of bias, then the pool of available arbitrators would be drastically narrowed. This would no doubt be problematic for the growth of arbitration – at least as it relates to Canadian construction law, particularly as the judiciary continues to work through the backlog of cases created by the COVID-19 pandemic.

As well, and as recognized by the Court in Aroma, arbitrators are not judges, and are remunerated by parties rather than the state; as a result, in our view, precedents applicable to the judiciary are not necessarily fully transposable to the arbitral context. Here the Court’s selection of the test appears to have subordinated the fact that, under the IBA Guidelines, this situation fell into the Orange category, perhaps resulting in a decision that appears inconsistent with the outcome in Halliburton despite their similar factual matrices.

Finally, and as noted above, this issue raises questions as to how courts should interpret the intent of counsel. It is plausible that rather than retaining an arbitrator a second time in order to curry favour, counsel might retain them on the basis that the arbitrator demonstrated a high level of proficiency in their role as arbitrator (competent case management, strong grasp of the issues, etc.).

Given the broad significance of the questions raised by Aroma, we look forward to seeing how it will be subsequently interpreted or applied whether again by the Ontario Superior Court of Justice or in the appellate courts. One thing is clear, however – arbitrators should manage their practices with a strong emphasis on continuous disclosure.

Learn More About Domestic and International Arbitration

Singleton Reynolds has a longstanding history in alternative dispute resolution, including construction and commercial arbitrations. Our experience in construction-related arbitrations, both domestic and international, is extensive.