Confidentiality versus transparency: reconcilable differences?

It’s fair to say that striking the right balance between arbitral confidentiality on the one hand, and transparency on the other, raises questions of legitimacy on both sides of the debate.

And it is often framed as involving tension between party autonomy on the one hand, and public interest on the other.  

But is that a fair characterisation?  And where should the balance be drawn, if it is?  Those are the points I want to explore this morning. 

We’ll start with the terms that will be discussed throughout the day. Confidentiality, is of course, simply the obligation of non-disclosure which binds all of the participants in the arbitration process.

No more, and no less than that.  We know that the concept of privacy is distinct, meaning simply that arbitration hearings are not open to the public – but of course, the two very often go hand in hand.

As for transparency, the dictionary definition is ‘the quality of being done in an open way without secrets’.[1] 

Well, there are a lot of secrets to keep in arbitration; for lawyers, keeping secrets is a basic part of the job description; but transparency is principally about justice in public, in the open. 

As Lord Justice Toulson memorably said in the Guardian v City of Westminster Magistrates case in the English Court of Appeal,[2] the point is that open justice “lets in the light and allows the public to scrutinise the workings of the law for better or for worse“.  

He also noted that publicity is the surest of all guards against impropriety, particularly for judges.  He framed this as a constitutional principle recognised by the common law since the fall of the Stuart dynasty – and at different points in this magnificent city’s history, that may have been an unfortunate choice of historical reference.

Lord Justice Toulson also said that fine words butter no parsnips, and the same is true about fine and high principles concerning the public administration of justice, particularly when what we are talking about is commercial arbitration – something which is emphatically not a public process, but a private form of dispute resolution.

So in a private form of dispute resolution, what sort of transparency are we really looking for? Do we value transparency of the substance – are we interested in the outcomes of particular arbitrations?  Or is transparency around the process more important?

This morning I’ll touch upon three main subjects.

1. Why the users of arbitration say that confidentiality matters to them.
2. What the confidentiality landscape looks like, and where Scotland fits into that picture.
3. And finally, the difficult balance between confidentiality and transparency, and my thoughts on where the dividing line should be drawn.

Now, different factors and interests apply to commercial arbitration in contrast to ISDS, of course.  In that context a whole series of factors, and in particular very pronounced public policy interests, naturally tend towards a more transparent approach – but I am going to focus on commercial arbitration this morning.  

Importance of confidentiality to users of arbitration

First, as to the importance of confidentiality to the users of arbitration. One of the most refreshing, I might say most enjoyable, aspects of arbitration is that it is a party-led, party-driven process.  That party autonomy is of course reflected in statute – we find it in section 1(b) of both the Scottish and the English Arbitration Acts, as also in Article 19(1) of the UNCITRAL model law – all of which provide that the parties should be free to agree how their disputes are resolved.

Confidentiality is undeniably an influential factor for parties in choosing arbitration over litigation.  Of course it’s not the only factor; but we can fairly say it’s an important one.   

As to that, in the 2018 Queen Mary Survey, 87% of respondents believed that confidentiality in international commercial arbitration is important.[3] Overall, confidentiality was rated the fifth most valuable characteristic of arbitration – so it’s not the be all and end all, but it is important, and it is valued.

So why is confidentiality important to the users of arbitration?  The reasons are straightforward, and easily understood. 

Their disputes may involve trade secrets, know-how, or other information that parties don’t want in the public domain.

And parties generally don’t want their disputes, sometimes embarrassing or involving reputational issues, aired in public.

And it’s not a glib observation to point out that even the stress of the experience, for witnesses and clients, can be less onerous in a private and confidential forum.

Confidentiality: An overview of the status quo

I turn to my second topic, and a short look at the landscape.

Whether an arbitration is confidential turns upon the parties’ agreement, the applicable laws, and the procedural rules. Let’s consider two sources: first, the arbitration itself and second, court proceedings relating to arbitration.

As to the arbitration itself, the position in Scotland is commendably clear and simple.

Confidentiality is on a statutory footing in Scotland. Rule 26 of the Arbitration (Scotland) Act 2010 is a default one, and it provides that the arbitrators and the parties must not disclose confidential information relating to the arbitration unless the parties agree, or one of the exceptions applies. And the disclosure of any confidential information relating to the arbitration is actionable as a breach of confidence.

The definition of “Confidential information” means information relating to the dispute, the arbitral proceedings, the award, or any civil proceedings relating to an arbitration where an anonymity order has been granted. So there is a very strong starting point in favour of confidentiality in this jurisdiction.

If we look south to the position in England and Wales, we will find a more nuanced position, you could say, although again, one which presents a heavy presumption of confidentiality. 

The Arbitration Act 1996 contains no provisions on confidentiality, unlike other national legislation.  And there’s nothing in the pending Arbitration Bill, either, about confidentiality, so we won’t see a change in the statutory landscape any time soon.   

Nevertheless, the default position is that arbitrations seated in England and Wales are governed by general principles of confidentiality. A duty of confidentiality is implied into the agreement to arbitrate. And that’s an approach which has been adopted in other jurisdictions, for example in Singapore.[4]

The exceptions to confidentiality have been developed through English case law, with the leading case on those exceptions still the Court of Appeal case of Emmott v Michael Wilson & Partners Ltd, more than 15 years ago now. This case law approach is effective, but does tend to produce a landscape pockmarked with specific exceptions, dug out over time.   

Broadly categorised, those exceptions are: 1) where there was express or implied consent; 2) where there is an order of the court (though the court does not have a general discretion to lift the obligation of confidentiality); 3) where it is reasonably necessary for the protection of the legitimate interests of a party to the arbitration; 4) where the interests of justice require; and, perhaps, (5) where public interest requires it.  There are some fairly wide discretionary tests at the back end of those exceptions, we can say.

Elsewhere, arbitration is not necessarily confidential by default – the UNCITRAL Model Law does not contain any provision on confidentiality, though of course that does not prevent party agreement. The position is the same in France, Switzerland and Germany, where the parties must expressly agree to keep their arbitrations confidential.

And then, at the other end of the spectrum, we can see jurisdictions like the USA, where the courts have refused to find a duty of confidentiality in the absence of an express contractual provision, or the adoption of a set of arbitration rules which provide for confidentiality.[5]

That accounts for confidentiality arising from the arbitration itself. 

The intersection of arbitration and court proceedings raises numerous issues. Not just through challenge applications, but also as to disclosure and arbitrator removal.

For example, where a party applies to set aside an award, the following issues are likely to arise.  First, whether the hearing should be conducted in private or public.  As a practical aside, at the last section 68 set aside application I did at the Rolls Building, this aspect was dealt with by a piece of A4 with the word ‘Private’ hastily scrawled upon it, and stuck to the court door with a bit of Sellotape, not long before the hearing.

Other considerations are the availability of court documents; whether the judgment can be published; and whether the judgment, if published, should be anonymised, or with another form of safeguard put in place?

Obviously the position varies from jurisdiction to jurisdiction, and I will limit myself to Scotland and England & Wales.

As to this, Scotland again takes a very clear approach that is set out in statute. The anonymisation of judgments is not automatic; but a party can apply to the court for an order prohibiting the disclosure of the identity of a party to the case.

If an application is made, by section 15(2) the Court “must grant the order” unless it is satisfied otherwise by reference to a short list of exceptions, most notably that publication would be in the public interest (section 15(2)(c)), or would be necessary in the interests of justice (section 15(2)(d)).   And we’ll come back to those concepts when we look at the position in England and Wales.

The key case on this in Scotland remains Arbitration Application No. 2 of 2011 which concerned the appeal of an arbitration award. The appeal was refused, and the award was upheld.

Lord Glennie, who is of course with us this morning, granted the application for anonymity. It was agreed at the end of the hearing that the Opinion should be issued to the parties in the first instance, to enable them to make representations as to whether there should be publication, and if so, whether any details could be omitted without removing from his decision “such sense as it might otherwise have“, as it was put, rather self-deprecatingly.

Neither party raised any difficulty about the Opinion being issued in this form. The tension between the statutory protection of anonymity and the public interest in open justice was thus recognised by the Court, and an opportunity given to the parties.  

I suggest this case provides a neat example of the Scottish courts upholding the statutory provisions concerning confidentiality and transparency. 

Again the position in England is a little more nuanced, but similar concepts of the public interest and the interests of justice apply in the jurisdiction.  There is no doubt that the involvement of the courts does fundamentally qualify the position on confidentiality. As Mr Justice Briggs stated in Milsom v Ablyazov: ‘…arbitration confidentiality or privacy is not absolute. Its preservation in any particular situation, for example, an arbitration appeal, is only the starting point and may be overridden where either the public interest or…the interests of justice require’.

And in particular, the courts have been clear that they will need to weigh the public interest in publication (especially if the judgment raises matters of general importance) against the parties’ expectation of confidentiality or privacy.

In the key English case on this, City of Moscow v Bankers Trust,[6] the Court of Appeal explained that when a party calls upon the court to exercise its supervisory role, the court is acting as a branch of the state.  It is not a mere extension of the consensual arbitral process.

From there, the starting point was that judgments on s68 should be made public, because there is legitimate interest in the operation and practice of arbitration.

But that need not mean revealing confidential information unnecessarily. The Court in City of Moscow v Bankers Trust was clear that it “did not consider that a party inviting the Court to protect evidently confidential information about a dispute must necessarily prove positive detriment beyond the undermining of its expectation that the subject matter would be confidential.” 

So we see that there is a starting point of a legitimate expectation of maintaining confidentiality around confidential information; you don’t need to prove ‘special damage’.

But again, that expectation can never be determinative.  We saw that earlier this year in the Commercial Court case of Mordchai Ganz v Petronz [2024] EWHC 101, where the Court decided to publish a non-anonymised judgment on an arbitration claim about challenges to an arbitration award. It was argued that there was a natural expectation of confidentiality given the confidentiality of the underlying LCIA arbitration; but the Court held that the expectation was insufficient, when balanced against the public interest benefits.

My personal view is that it is difficult to see why the court felt that the public interest arguments outweighed the expectation of confidentiality in this case, given that it was a private dispute about a share purchase agreement, but it underlines that once you involve the English Court, you do so at the price of opening the door to the Court’s assessment of public policy and the interests of justice – and don’t be surprised if you lose the benefit of your arbitral confidentiality regime.

The dilemma: confidentiality versus transparency

We can see from this brief review of the cases that in Scotland, as also in England & Wales, and many other jurisdictions besides, there is a tension between the parties’ legitimate expectation of confidentiality and their public interest in the administration of justice.

What is that public interest? And what are the arguments for transparency?

Well, you really have to start with the principle of open justice. Public scrutiny is desirable in itself, because it’s a means by which confidence in the courts can be maintained, and the administration of justice can be made transparent.[7]  And the Supreme Court underlined that in 2019 in the Cape Intermediate Holdings case.

Another argument is that transparency bolsters the legitimacy of arbitration as a dispute resolution method – because it leads to more certainty, more predictability. and as I mentioned before, there is a legitimate interest in its operation and practice.[8]  

It’s often also said that transparency aids the development of arbitration jurisprudence, and more broadly still, the development of the common law – even if awards are not binding, they may have persuasive value, and they may be cited later to develop the law by reference to them.  And finally, you can argue that an expectation of public scrutiny may improve the quality of arbitration awards.

How well do these arguments hold up, that’s the question.

The “open justice” argument is less persuasive, I suggest.  Arbitration is a private form of dispute resolution arbitration distinct from the court process, paid for and agreed by the parties. It’s a very different forum with different legitimate expectations.

Speaking of which, the legitimacy – perceived or actual – of the arbitration process is a more relevant consideration than open justice, to my mind. However, there is a good counter-argument to that, which is that a key element of legitimacy and trust in the system is, in fact, its confidentiality.

On that, Sir Patrick Neill QC said in his 1995 “Bernstein” Lecture that it would be difficult to conceive of any greater threat to the success of arbitration, in England at least, than the removal of the general principles of confidentiality and privacy.

I wouldn’t go that far.  I can think of worse things to happen to arbitration as a process.  But still, it’s a powerful point. We must be careful not to disregard confidentiality in the quest for transparency. Where an arbitration runs without the need for the courts to supervise its process or enforce its outcome, there can be no compelling justification for the parties to share any aspect of these proceedings publicly unless they choose to do so. We may be interested to read all about it; but we’ve no real right to do so. 

As for the development of the law and jurisprudence, the time is short to go into that very interesting, sometimes divisive topic, one that’s been much debated in recent years.  But only to say that to my mind the starting point is that it’s for the arbitrators to apply the law, rather than make it.

If points of law arise in a case, in some jurisdictions the parties retain the right to appeal to the Court on them, and that generates new law on sometimes interesting points.  But it’s instructive to see how often, in practice, parties choose to waive that right and to prioritise finality and speed of result, when it’s them who are paying for the privilege of arbitration.  And I’m of the view that the incremental development of the law is a very heavy responsibility to lay at the door of the users of arbitration.

But there are still ways of increasing transparency without undermining the confidentiality which is valued highly by the users of arbitration.  There are “reconcilable differences”, to use that phrase.

First, it’s possible to increase transparency, and confidence in the process, through the publication of anonymised awards.

There are far more commercial awards available than ever before on platforms like Kluwer and Jus Mundi. Provided the parties don’t opt out, the ICC now publishes anonymised awards through these platforms. Jus Mundi’s website now holds around 4800 final awards; Kluwer Arbitration has more than 7,500.

Some institutions publish decisions directly on their websites, with party agreement – notable examples are the LCIA and HKIAC.

But of course, institutions are careful to respect party choice – the push for transparency must not override what users want, and that’s why the consent of the parties is needed – whether by opt-in, or opt-out sysytems.

Second, it is even more straightforward to provide for transparency in the system of arbitration. By this I mean how arbitration works, and how institutions administer cases, through the publication of caseload statistics, procedural decisions, statistics on arbitrator nationality, and gender diversity, and other critical data points of real interest. That is not confidential to parties. It is procedural, not substantive.

Potential users can benefit from the sharing of aggregated data. And the publication of this information can enhance the legitimacy of the system of arbitration, without compromising on party confidentiality.

I conclude with this. 

Arbitration is a very special thing.  I will try not to be overcome with emotion, which is a very Scottish trait when you talk about things that you love.  Honestly, arbitration is a phenomenal thing, really, when you stop to reflect upon it.  It is an incredibly effective system of dispute resolution; one that offers an extraordinary, unmatched transnational reach. 

But it is also a generally private system, too, and we must never lose sight of that.  It is party-led, and party-funded; and the party expectation of confidentiality is routinely assessed by the users themselves as one of its most attractive features. 

I am desperately sorry that I’ll miss this evening’s ceilidh, but as will be seen later on, he who pays the piper calls the tune.  So must it also be in arbitration – with narrow and carefully prescribed exceptions only. 

Because the quest for transparency is understandable, and it is sometimes laudable.  But we cannot let the pursuit of high principle put the attractiveness or legitimacy of commercial arbitration at risk. 

Thank you very much.

[1]        https://www.kluwerarbitration.com/document/kli-ka-bull-2023ch14?q=transparency%20in%20commercial%20arbitration

[2]        R (Guardian News and Media Ltd) v City of Westminster Magistrates Court [2013] QB 617.

[3]        QMUL Survey 2018, The Evolution of Arbitration, p3.

[4]        Myanma Yaung Chi oo Co v Win Win Nu [2003] SLR. 547 (Singapore High Court)

[5]        U.S. v. Panhandle et al. (1988) 118 F.R.D. 346 (D. Del) —in United States. In 1988, a United States federal district court found that, absent explicit agreement by the parties or institutional rules on point, arbitration proceedings are not necessarily confidential. In Panhandle Eastern, the United States government sought the production of documents related to a previous arbitral proceeding in Geneva, held under International Chamber of Commerce (“ICC”) rules. The Court ruled that because the arbitration agreement and applicable arbitration rules did not provide for the confidentiality of the proceedings, the government could access the documents. Fundamentally, the Court failed to recognize any general principle of confidentiality in international arbitration, and under this holding, any duty of confidentiality can only be implied-in-fact. The import of this holding, at least in the United States, was to underscore the necessity of broad confidentiality clauses in arbitration agreements.

[6] [2004] EWCA Civ 314

[7] The Supreme Court’s decision in Cape Intermediate Holdings Ltd v Dring [2019] UKSC 38 identified the two principal purposes of the open justice principle, namely:

      to enable public scrutiny of the way in which courts decide cases, holding judges to account for the decisions they make and enabling the public to have confidence that they are doing their job properly; and

      to enable the public to understand how the justice system works and why decisions are taken, by ensuring they are in a position to understand the issues and the evidence adduced, particularly in light of modern court practices which mean that much of the argument and evidence is in written form rather than being presented orally at the public hearing.

[8]        Mordchai Ganz v Petronz [2024] EWHC 1011: https://hsfnotes.com/arbitration/2024/05/21/english-commercial-court-rules-in-favour-of-publishing-a-judgment-on-an-arbitration-claim/