Image Alt

The Singapore Law Gazette

Acting Against the Same Opponent in Similar Matters

Rebalancing Client Choice and Confidentiality

This article examines the competing policy considerations involved where a lawyer acts for different clients against the same opponent in similar claims. With reference to Singapore, Australia, New Zealand and English cases, it suggests that the interests of client choice and the protection of an adversary’s confidentiality should be rebalanced in Singapore.


Should a lawyer be disqualified from acting for a client merely because the lawyer had acted for a previous client against the same opponent in a similar matter for which that opponent had disclosed confidential information to the previous client and the lawyer during settlement discussions? Rare as this scenario may seem, the issue has been considered in the past two decades by a number of common law jurisdictions, including Singapore, Australia and New Zealand. The July 2019 decision by the High Court of England and Wales in Glencairn IP Holdings Limited and another v Product Specialities Inc (t/a Final Touch) and another1(2019) EWHC 1733 (IPEC). (“Glencairn”) is the latest instalment of this string of multi-jurisdictional cases that have been difficult to reconcile.

The policy arguments on both sides, although not always expressly articulated in every case, are significant. On the one hand, it is desirable that litigants should not be deprived of counsel of their choice.2See e.g. British Sky Broadcasting Group plc and another v Virgin Media Communications Limited and others (2008) EWCA Civ 612 at (20). Moreover, a solicitor’s duty to continue to represent a new client should not be controlled by his adversary.3Williamson v Schmidt (1998) 2 Qd R 317 at 318 and 326. On the other hand, “there is an inherent incompatibility between lawyers’ participation in a confidential mediation and their desire to act for other clients in parallel litigation” as “a lawyer cannot compartmentalize his or her mind”.4Carter Holt Harvey Forests Ltd v Sunnex Logging Ltd (2001) 3 NZLR 343 at (30).

To a significant extent, these competing policy concerns are reflected in how a court formulates the legal test to resolve the issue, and how that test is applied to the subtle factual nuances of each case. This article will first examine the approach taken in Glencairn (which may be going up on appeal)5See e.g. <> (accessed 10 February 2020). before turning to the position in Singapore. It then discusses, from the Singapore perspective, how the interests of client choice and the protection of an adversary’s confidentiality should be rebalanced.

Glencairn and the Partial Bolkiah Test

In discussing Glencairn and similar cases, the following typical basic factual scenario will be used in this article:

  • A lawyer or law practice (L) acted for a previous client (C1) against an opponent (O) in a matter (M1), which is successfully mediated or settled through negotiations;
  • O’s confidential information was disclosed to L and C1 during the M1 mediation or negotiations;
  • L’s confidentiality obligation to O arose from either an express condition of confidentiality or in circumstances where a duty of confidence is implied; and
  • L then acts for a new client (C2) against the same opponent O in a similar matter (M2).

In Glencairn, L, a law practice with an office each in two separate UK cities, acted for C1 in registered design infringement proceedings (M1) brought by O. C1 was represented by a small team of solicitors from L (C1’s team). The M1 claim was successfully mediated, under a mediation agreement which expressly bound the parties and the solicitors present to keep confidential information disclosed at the mediation. Subsequently, the parties entered into a settlement agreement.

By the time of the M1 mediation, L had taken the view that C1’s team should not be involved in a similar action brought by O against another party, C2, for similar infringement (M2). An information barrier was then set up between C1’s team and another small team of solicitors from L acting for C2 (C2’s team). C1’s team informed O’s solicitors that they would not be acting for C2 in M2 and that an information barrier had been implemented. However, O’s solicitors were not impressed and informed L of their doubts that any information barrier set up between C1’s and C2’s teams could be effective. Subsequently, O applied for an injunction to restrain L from acting further for C2 in M2.

The basis of O’s application was that L (via C1’s team) became aware of information disclosed by O during the M1 mediation and during discussions both leading up to it and afterwards, including O’s negotiating position and the terms on which O was prepared to settle. Hence, there was a risk that C2 would be privy to such information (via L/C2’s team) and would have an advantage in M2, particularly in any ensuing settlement negotiations.

A threshold question that Judge Hacon in the High Court of England and Wales considered was whether the principles established in Prince Jefri Bolkiah v KPMG6(1999) 2 AC 222. (Bolkiah) applied to Glencairn. Bolkiah concerned an injunction sought by a former client to restrain his professional advisers (forensic accountants) from acting against him in a subsequent matter on the basis that the court should intervene to protect the former client’s confidential information. In this landmark decision, which also applied to lawyers, the House of Lords established the well-known Bolkiah test in determining whether an injunction should be granted:

  • The former client must first establish: (a) that the lawyer is in possession of information which is confidential to him and to the disclosure of which he has not consented; and (b) that information is or may be relevant to the new matter in which the interest of the other client is or may be adverse to his own; and
  • Once the former client establishes both (a) and (b), the evidential burden shifts to the law firm to show that even so there is no risk that the information will come into the possession of those now acting for the other party.

Judge Hacon noted that what fundamentally distinguished Glencairn from Bolkiah was that L had never acted for O. Therefore, no fiduciary relationship arose between L and O and the Bolkiah test “should not be applied with full force”7Supra, n 1 at (45). to the facts of Glencairn.

In examining the authorities, the court identified three classes of cases. The first class of cases comprised actions like Bolkiah where a former client sought to injunct a solicitor from representing a party with an interest adverse to the former client. Tighter safeguards against wrongful disclosure of information were “proportionate in the Bolkiah type of case”8Id, at (40). because of the additional risk of “a breach of the continuing fiduciary duty owed by a solicitor to a former client”.9Id, at (22). Glencairn obviously was not such a case, but it also did not fall into a second class of cases where the risk of disclosure of confidential information obtained from an adverse party was “the ulterior use”10Id, at (41). of documents obtained through discovery. Instead, cases like Glencairn belonged to a “third, intermediate class of cases”,11Id, at (43). where the overall burden of proof remained with O and would not shift to L, unlike in the original Bolkiah test.12Id, at (52)-(53). As such, it was for O to prove that L’s information barrier was not effective.

On the facts, Judge Hacon declined to grant the final injunction for two reasons. Firstly, he found only a low likelihood of any confidential information being passed to C2. Here, a significant issue was the effectiveness of L’s information barrier, which in turn raised the question of whether the risk of disclosure across an information barrier was more, or less, likely in a small firm. This was because L was considered as a small firm in the UK context. Judge Hacon assumed that there was a higher risk of an information barrier in a small firm being “less developed and therefore less effective than one in a large firm”.13Id, at (86(a)). But he did not conclude that “a disclosure barrier in a small firm [could] never be effective” as this was fact-dependent.14Ibid.

Judge Hacon observed that although it was inevitable that there would be regular contact between C1’s and C2’s teams in a small firm like L, it was also likely that all members of both teams were “highly aware that nothing should be said about the [M1] litigation.”15Id, at (86(e)). In addition, L had restricted the access of the M1 documents to C1’s team only and each team worked in separate offices. Despite the lack of detailed evidence on how L’s information barrier worked, Judge Hacon was satisfied that its supervision by the lead lawyer in C2’s team sufficed without the need to institute formal monitoring and disciplinary arrangements as suggested in Bolkiah.

Secondly, Judge Hacon held that it was “extremely unlikely, to the point of being fanciful”16Id, at (94). that the entire settlement agreement (which was not disclosed to the court or to C2’s counsel in the application) would be disclosed to C2, such that O would be prejudiced in any negotiations with C2. Conversely, C2 would likely be prejudiced if the injunction was granted, as he assumed that C2 had instructed L “for good reason” and “a good working relationship [had] developed” between C2 and C2’s team.17Id, at (93). Client choice was therefore a relevant, though not determinative, factor.

So is Glencairn representative of the case law jurisprudence in this area, in terms of balancing the competing policy interests? No, Glencairn is fact an outlier because its facts concerned separate matter teams within L acting for different clients. In contrast, the precedent cases that Judge Hacon cited and discussed in his judgment all involved a single lawyer or one set of lawyers, which triggered the inescapable inference that the lawyer(s) cannot build an information barrier in his or her mind. We next turn to the Singapore position to examine a recent case where only a single lawyer was involved in acting for different clients.

The Complicated Position in Singapore

In an April 2019 decision by the Singapore High Court (which pre-dated Glencairn), Wan Hoe Keet and another v LVM Law Chambers LLC18(2019) SGHC 103. (Wan Hoe Keet), the Judge was presented with facts resembling Glencairn, but with three notable factual differences.

Firstly, Wan Hoe Keet did not involve different teams of solicitors representing different clients, but only a single lawyer. Secondly, the M1 claim in Wan Hoe Keet was settled via negotiations, not mediation, and the settlement agreement did not expressly impose a contractual duty of confidence on L. Thirdly, the similar M2 claim in Wan Hoe Keet contained an additional allegation against O (in this case, two persons) that was not made in the M1 claim.

On the first issue of whether L owed a duty of confidentiality to O, the High Court referred to the 2001 New Zealand Court of Appeal decision of Carter Holt Harvey Forests Ltd v Sunnex Logging Ltd19Supra, n 4. (Carter Holt) and the 2009 New South Wales Court of Appeal decision in Worth Recycling Pty Ltd v Waste Recycling and Processing Pty Ltd20(2009) NSWCA 354. (Worth Recycling). Specifically, the High Court pointed out that the facts of Worth Recycling and Wan Hoe Keet were substantially similar.

In Worth Recycling, although M1 had been settled by a mediation, L was not contractually bound by the terms of the mediation agreement to preserve O’s confidentiality as L was not a party to the agreement. Nevertheless, the Court found that L owed an obligation of confidence directly to O in view of the circumstances in which the confidential information was imparted to L. This was because L knew that C1 had promised O that L would not use confidential information except in accordance with the mediation agreement, and L participated in the mediation in that knowledge.

Applying Worth Recycling, the High Court similarly found in Wan Hoe Keet that L was bound by an implied duty of confidence. It observed that L could owe a duty of confidence to O in equity “by applying principles of good faith and conscience”, even though no contractual relationship existed.21Supra, n 18 at (9). An equitable duty of confidence would be imposed if the circumstances were such that a reasonable solicitor in L’s position should have known that the information was given in confidence.

On the facts, the settlement agreement required the parties to keep the settlement (including its terms) strictly confidential. Even though the clause did not expressly impose a contractual duty of confidence on L, it was clear that L knew that his client C1 promised O that he would not use or disclose any of these confidential information except where contractually provided. Hence, “an equitable duty of confidence” was imposed on L “to not divulge or use the confidential information obtained from the negotiations except” in accordance with the M1 settlement.22Ibid.

The High Court further observed that “[t]he fact that parties negotiated instead of mediated” was irrelevant because like the solicitors in Worth Recycling, “[t]he private and confidential nature of the negotiations create the same nature and degree of fidelity and L, as solicitor, is bound to the confidential agreement that his client [C1] signed with [O]”.23Ibid.

The High Court also agreed with counsel’s submission that O’s information was confidential because the settled amount and the terms of a settlement, as well as “the nature and process in which that settlement sum was reached” were important confidential terms.24Id, at (7).

On the second issue of whether there was a threat of misuse sufficient to justify an injunction, the High Court held that notwithstanding L’s assurances that he would not be thinking about the settlement sum or divulge it to C2, the concern was the possibility of “a future breach occurring accidentally or unconsciously” due to “subconscious currents in our minds”.25Id, at (10). In this regard, the High Court cited Hodgson JA’s conclusion in Worth Recycling that misuse was “almost inevitable if [L] should take part in any settlement negotiations”, especially in light of Carter Holt’s observation that it “is very difficult to keep the settlement negotiations quarantined from the conduct of the proceedings generally”. 26Ibid.

The High Court noted that it was likely that C2’s action might also involve negotiation, and regardless of whether the negotiation succeeded or not, O would be disadvantaged from “the knowledge [L] possesses from his participation in the M1 negotiations, just as C2 would “gain an advantage of inside knowledge he would otherwise not have”.27Id, at (11). The High Court added that L “will know at which point [O] became malleable and at points they are at their strongest”.28Ibid.

An intriguing question raised by Wan Hoe Keet is whether the High Court had in fact impliedly approved or applied the “real and sensible possibility of misuse” test used in Australian authorities such as Worth Recycling. Although the High Court followed Worth Recycling, it did not expressly refer to this test in its judgment. This is a significant point because subsequent Australian case law indicates that mere possession by L of O’s confidential information is insufficient and it is critical to make a finding of “a real and sensible risk of misuse” of O’s confidential information.

In the 2011 decision of Ian West Indoor and Outdoor Services Pty Ltd v Australian Posters Pty Ltd29(2011) VSC 287. (Ian West), which was not cited in either Wan Hoe Keet or Glencairn, the Supreme Court of Victoria held that it is not enough to show that L “could not expunge from his mind the knowledge of the confidential negotiations in the [M1] proceeding”.30Id, at (49). The Court must go further to consider whether there is a real and sensible risk of misuse by L of confidential information “concerning the settlement amount, the steps by which settlement was reached and the movements in negotiations in the [M1] proceeding”.31Ibid. As the burden of proof remained with O, an appearance of risk “going beyond the remote or merely fanciful” was insufficient.32Id, at (48). Neither could the onus of proof be shifted to L simply because “a common factual element” existed for both matters.33Ibid.

Even if the “real and sensible possibility of misuse” test had indeed been adopted in Wan Hoe Keet, the position in Singapore is complicated by a prior decision issued by another Singapore High Court judge in July 2018. In Harsha Rajkumar Mirpuri (Mrs) née Subita Shewakram Samtani v Shanti Shewakram Samtani Mrs Shanti Haresh Chugani34(2018) 5 SLR 894. (“Harsha”), the High Court was faced with an analogous scenario where lawyers owed an equitable duty of confidentiality to a former prospective client. In holding that no fiduciary relationship existed in this scenario, the High Court adopted a partial Bolkiah test similar to that in Glencairn:

“… where the plaintiff is merely a former prospective client, he should not obtain the advantage conferred by that part of the Bolkiah test that places the burden on the law firm to prove the absence of a risk of disclosure.”35Id, at (47).

Given that both Harsha and Wan Hoe Keet were cases where no fiduciary duty was held to exist between a lawyer and his former prospective client or adversary, it would not appear consistent, in light of Glencairn’s classification approach, for a court to apply, in a future case, the partial Bolkiah test vis-à-vis a former prospective client and the Australian “real and sensible possibility of misuse” test vis-à-vis an adversary. A choice will have to be made as to which approach to adopt for both scenarios.

Rebalancing Client Choice and Confidentiality – the Singapore Perspective

As it stands, Wan Hoe Keet is the only Singapore authority on a lawyer’s duty of confidentiality where the lawyer acts against the same adversary in similar matters. Glencairn provides a useful prism to reassess the approach taken in Wan Hoe Keet, even though the partial Bolkiah test was not adopted in the latter case. Harsha also provides a necessary lens as to how the partial Bolkiah test, if adopted in a future Singapore case, could be applied to an adversary scenario. It is submitted that regardless of whether the partial Bolkiah test or the “real and sensible possibility of misuse” is adopted in Singapore in future, a rebalancing of client choice and confidentiality is essential. To ensure that the importance of client choice is given sufficient weight, the Singapore courts should consider employing a number of conceptual or analytical tools to assist in their evaluation.

Firstly, the Court should be satisfied that the two matters M1 and M2 are not merely similar, but also relevantly similar. In Ian West, the Supreme Court of Victoria did not grant an injunction as it was also not persuaded on the facts that the two proceedings were “so relevantly similar that knowledge of offers and negotiations in the [M1] proceeding will give rise to a real and sensible possibility of misuse of that information in the [M2] proceeding”36Supra, n 29 at (54). [emphasis added]. This would entail not only requiring O (who bears the burden of proof) to identify and analyse the similarities between M1 and M2,37Id, at (52). but also to prove that the issues and evidence in M1 and M2 are not likely to be so different that knowledge of the confidential information disclosed by O in M1 would be “of negligible use or relevance” in M2.38Tricontinental Corporation Ltd v Holding Redlich (a firm) (1994) VicSC 829 at p 9. Superficial similarities in pleadings for similar proceedings should be not be the yardstick. Likewise, claims that L would simply be able to “assess and transpose” knowledge of O’s confidential information from M1 to M2 should be carefully scrutinised.39Id, at p 8.

Secondly, it should not be assumed that L’s knowledge of O’s negotiating moves and settlement amounts automatically translate to an advantage for C2 in M2. Certain threshold questions need to be addressed in the first place, for example:

  • Why is it assumed that O would employ the same negotiating moves in M2? Such an assumption would have to be premised ultimately on whether the two matters are relevantly similar (see above). Even if such an assumption is correctly premised, it should not be determinative given that O is not obliged in any event to use the same negotiating tactics in M2. Giving too much credence to this assumption would effectively allow O to control L’s representation of C2.
  • How might L misuse the information concerning the settlement amount in M2 for C2’s benefit? As noted in Ian West, the court must consider whether O’s confidential information might be “capable of being misused in the [M2] proceeding, that is applied for the benefit of [C2] to the disadvantage of [O]”.40Supra, n 29 at (49). An English commentator observed that Carter Holt and Worth Recycling appeared to have “taken a very generous view” as to the “likely importance” of O’s confidential information (such as the settlement amount) and doubted whether such information was “genuinely relevant” to M2.41Charles Hollander QC & Simon Salzedo QC, Conflicts of Interest (5th edition, United Kingdom: Thomson Reuters (Professional) UK Limited, 2016), at (9-013).
  • Did C2 engage L with the intention of taking advantage of L’s knowledge of O’s confidential information in M1? In Ian West, the court distinguished Worth Recycling on the basis that C2 had chosen to retain L with the expectation of obtaining a favourable settlement agreement on terms similar to those obtained by C1. In contrast, in Ian West, C2 had retained L as his solicitor before C1. Hence, it could not be suggested that C2’s motive was to take advantage of L’s knowledge of the M1 proceeding. Moreover, Glencairn had also recognised that a bona fide reason for C2’s retention of L in M2 would be a factor in supporting a decision to refuse the injunction.

Thirdly, if the Singapore courts adopt the partial Bolkiah test in an adversary scenario in future, client choice should be recognised as an important policy consideration. In Glencairn, Judge Hacon held that although the Bolkiah test did not involve a balancing exercise, the impact of a Bolkiah order on the current client should be taken into account given that was “no relevant fiduciary relationship” between the lawyer and his or her adversary.42Supra, n 1 at (51). In Harsha, however, the High Court did not expressly refer to client choice as a policy consideration when it considered whether to restrain the law practice from acting under the law on breach of confidence.43Client choice was only referred to in the High Court’s discussion on whether the law practice should be restrained from acting under the court’s supervisory jurisdiction: supra, n 34 at (82). The position in Harsha on this point should be reassessed, especially since it is an immutable rule that no fiduciary relationship can exist between a lawyer and his or her adversary.


In 2001, the New Zealand Court of Appeal in Carter Holt dismissed fears of “frequent disqualification applications” and deprivation of client choice as “greatly exaggerated” because it would only be “in a relatively small number of cases” where a client would employ the same lawyer or law practice to act against the same adversary in a similar matter.44Supra, n 4 at (29). With the increasing emphasis on mediation and other forms of alternative dispute resolution in the past decade, things look very different in 2020 and there is a real possibility that disqualification applications will become increasingly prevalent as a tactical weapon by shrewd adversaries to diminish client choice. As argued in this article, a rebalancing of the policy considerations in Singapore, through a more critical analysis and application of the applicable legal test, is necessary to recognise the importance of client choice, as well as to minimise the possible unintended consequence of a deteriorating incentive to mediate disputes.45Supra, n 41 at (9-014).


1 (2019) EWHC 1733 (IPEC).
2 See e.g. British Sky Broadcasting Group plc and another v Virgin Media Communications Limited and others (2008) EWCA Civ 612 at (20).
3 Williamson v Schmidt (1998) 2 Qd R 317 at 318 and 326.
4 Carter Holt Harvey Forests Ltd v Sunnex Logging Ltd (2001) 3 NZLR 343 at (30).
5 See e.g. <> (accessed 10 February 2020).
6 (1999) 2 AC 222.
7 Supra, n 1 at (45).
8 Id, at (40).
9 Id, at (22).
10 Id, at (41).
11 Id, at (43).
12 Id, at (52)-(53).
13 Id, at (86(a)).
14 Ibid.
15 Id, at (86(e)).
16 Id, at (94).
17 Id, at (93).
18 (2019) SGHC 103.
19 Supra, n 4.
20 (2009) NSWCA 354.
21 Supra, n 18 at (9).
22 Ibid.
23 Ibid.
24 Id, at (7).
25 Id, at (10).
26 Ibid.
27 Id, at (11).
28 Ibid.
29 (2011) VSC 287.
30 Id, at (49).
31 Ibid.
32 Id, at (48).
33 Ibid.
34 (2018) 5 SLR 894.
35 Id, at (47).
36 Supra, n 29 at (54).
37 Id, at (52).
38 Tricontinental Corporation Ltd v Holding Redlich (a firm) (1994) VicSC 829 at p 9.
39 Id, at p 8.
40 Supra, n 29 at (49).
41 Charles Hollander QC & Simon Salzedo QC, Conflicts of Interest (5th edition, United Kingdom: Thomson Reuters (Professional) UK Limited, 2016), at (9-013).
42 Supra, n 1 at (51).
43 Client choice was only referred to in the High Court’s discussion on whether the law practice should be restrained from acting under the court’s supervisory jurisdiction: supra, n 34 at (82).
44 Supra, n 4 at (29).
45 Supra, n 41 at (9-014).

Chief Legal Officer and Director, Representation and Law Reform
The Law Society of Singapore
E-mail: [email protected]