1. INTRODUCTION

The phenomenon of e-discovery has contributed to the spiralling costs of litigation in the United States, as well as in other jurisdictions, predominantly common law, where broad document production rules apply. The existence of electronically stored information ("ESI") in ever greater quantities is something that cannot be ignored by arbitration practitioners. Indeed, it has recently received considerable attention in the international arbitration press, as practitioners question the role the production of electronic evidence can, and should, play in the arbitral process. 1

The title of this paper is phrased as a single question, but in reality it is two separate questions with potentially different answers. Firstly, will electronic evidence change the face of arbitration? Secondly, will e-discovery change the face of arbitration? In fact, should e-discovery have any place in an international arbitration at all?

Information stored in an electronic format-i.e. electronic evidence-is different from paper documentation in several significant ways. This fact has led the courts in various jurisdictions to the view that specific rules or guidance are needed to deal with it. The attention given to the issue by litigation lawyers has also given rise to a wide-ranging debate in the international arbitration community as to whether rules or guidelines are appropriate in the very different context that is international arbitration. However, it goes without saying that the mere existence of electronic evidence does not in itself necessitate a common-law-style discovery or disclosure process in the arbitration context. 2[Page374:]

The right to disclosure or discovery recognized by the common law courts is an ancient one. It has been described as "an exercise of sovereign authority to require citizens and foreigners within the jurisdiction to assist in the administration of justice". 3 In other words, courts have the right to order a party to proceedings (or even, in some circumstances, a third party) to produce documents belonging to it (even internal and confidential documents) in order to shed light on the dispute that is before the court and assist the court in deciding it. As a result of the recognition by the common law of this right over the documentation held by parties to litigation, the rules of court in common law jurisdictions will typically provide that a party is required to hand over to its opponent broad classes of documents that are likely to assist in resolving the dispute between the parties (sometimes even before filing the brief on the merits).

This approach to the exchange of evidence is in sharp contrast with the civil law tradition, where parties are primarily only required to disclose, annexed to their written arguments, the documentation on which they intend to rely to prove their own case. 4 This is explained in terms of the right of each party to understand the case it has to meet, with sufficient time to allow it to prepare a response. Secondarily, civil courts now typically have rules allowing them to order the exchange of individual documents, or sometimes a very limited class of documents, that have been specifically requested by the opposing party. These orders to produce documents are rare and bear no comparison to what would be ordered by a common law court.

Arbitrators and parties in international arbitration will generally adopt some form of pragmatic compromise between these two approaches to document production, which will usually involve two phases:

1. First, parties produce the documents on which they intend to rely to satisfy their burden of proof.

2. Secondly, each party has the opportunity to request production of additional specific documents (or classes of documents) relevant and material to the outcome of the case.

When ordering production of documents, the tribunal is usually guided by the principles of relevance and proportionality, and must therefore balance the cost of production against its benefit for the outcome of the case, bearing in mind the amount in dispute. 5[Page375:]

Various guidelines exist which together demonstrate an evolution towards a loose consensus view on the production of documents, while still maintaining a degree of flexibility that allows orders for production to be tailored to the circumstances of the case. Perhaps the best known of these is the International Bar Association's Rules on the Taking of Evidence in International Commercial Arbitration, the second edition of which was adopted in 1999 ("the IBA Rules"). These rules trace a middle path between the common law and civil law approaches, by requiring the parties to submit all documents on which they rely and then allowing them to request each other to produce a "narrow and specific" category of additional documents. 6

The approach that international arbitration adopts to the phenomenon of ESI therefore cannot be an exact mirror of the one adopted in the rules of court of the United States, England or elsewhere. Far from it, international arbitration will need to continue to develop its own "best practice" in order to ensure it retains the flexibility of approach that distinguishes it from litigation.

It would not make sense for the existence of electronic evidence to push arbitration practitioners towards a more American approach to disclosure. At the same time, an informed debate about the subject of ESI in international arbitration is helpful and even necessary. The tidal wave of electronic information generated by the modern business has the potential, if not to change the face of arbitration, then at least to catch unwary parties by surprise and lead to unjust results in individual cases. After dealing briefly with the specificities of ESI and the changes that it has prompted in the rules of court in various jurisdictions and, subsequently, in arbitration guidelines, this paper sets out some considerations that will help to keep the issue in its proper perspective.

2. THE NATURE OF THE PROBLEM

a. Is ESI not just the same material in a different format?

Orders for disclosure, discovery or the production of documents are a means to an end: they allow a party to an action access to information held by another party in order to assist it in proving its case. Looked at in this light, the recent debate about e-discovery may seem like something of a storm in a teacup. Surely electronic evidence is simply the same information, stored in a different way? [Page376:]

Broadly speaking, of course, it is true that the information stored on electronic media in today's business world is often the same as or similar to information that would have been held in paper form 30, 20 or even 10 years ago.

The initial reaction of courts and arbitrators was therefore to recognize, from the early days of the "information revolution", an equivalence between paper and electronic "documents" and to underline that the mere fact that information is stored in such a way that a device is necessary to be able to decipher it-be it a microfilm reader, a CD player or a computer with appropriate software-does not exempt that information from the normal discovery, disclosure or document production regime. 7

However, there are particularities of the electronic media on which ESI is stored that have a significant impact on the document production process and mean that such an undifferentiated approach can lead to misunderstandings and complications. 8

b. Characteristics of ESI

An analysis of what makes ESI difficult to handle in the common law discovery process has been carried out by the Sedona Conference, a non-profit law and policy think-tank that has been influential in shaping the US debate about e-discovery. In its seminal publication, the Sedona Principles, 9 the Sedona Conference identifies six distinguishing features of ESI:

Volume and duplicability-ESI is generated in significantly greater quantities than paper documentation and is subject to rapid and large-scale user-created and automated replication without degradation of the data.

Persistence-ESI is more difficult to dispose of than paper documents, as "deleted" documents can often be retrieved.

Dynamic, changeable content-ESI is more easily and more thoroughly modifiable than paper documents, and changes can result from automatic processes even without direct human intervention.

Metadata-it is often possible to access data associated with an electronic document (which is not immediately visible to the viewer) to find out information about how that document was created, handled, transferred or stored. [Page377:]

Environment-dependence and obsolescence-because ESI requires a computer and software in order to be read, it is apt to become unreadable when data is moved to a different platform.

Dispersion and searchability-depending on the circumstances, ESI may be more difficult to search for than paper documents, because it may reside in numerous different locations, or easier, because of the existence of search software.

The most striking problem created by the electronic storage of information is that its volume is exponentially greater than that of paper documents. In a modern business, electronic information may be found on a mainframe, servers, networked workstations, desktops and laptops, home computers, removable media (such as CDs, DVDs and USB drives) and handheld devices (such as PDAs, mobile phones, iPods and so on). Each of these devices has a capacity to store massive amounts of information, and the ease with which information can be created and stored means that this capacity is inevitably used. The volume of information stored electronically will only increase with time, given the steadily decreasing costs and increasing capacity of computer and server storage space and the increasing reliance of businesses on electronic means of communication and document generation.

As well as leading to the creation of larger numbers of documents, modern technologies also encourage the indiscriminate duplication of information. For example, e-mails can be sent to multiple recipients, replied to, forwarded, received on multiple devices and so on. Documents may be saved on multiple devices and may be worked on in multiple versions by multiple users. Businesses will also typically have a routine system to back up their electronically held information and may well do so on a daily, weekly and monthly basis (with back-up records being overwritten after a prescribed period of time).

The task of searching for electronic information relevant to a dispute is complicated by the fact that there is most often no centralized point of control for the storage of data-each employee will typically hold a unique set of records that will overlap with, but not be identical to, that of other employees. [Page378:]

At the same time, the availability of search tools can mean that ESI can be cheaper to disclose than paper data (assuming the data is readily accessible and amenable to being searched). The preponderance of costs involved therefore relate to the need for a review of the documents to be carried out-by both the producing and the receiving party. In the American context, the Sedona Conference estimated the cost of lawyer review of a typical gigabyte of ESI (around 80 000 pages) to be in excess of US$30 000 (at 2007 rates). 10

c. Recent developments in the litigation world

The effects of ESI on the document production process were felt most strongly in jurisdictions with the broadest document production regimes, and it therefore follows a certain logic that, to date, the jurisdiction to implement the most comprehensive amendment to its rules of court to deal with ESI has been the United States. On 1 December 2006, amendments to the US Federal Rules of Civil Procedure ("the Federal Rules") came into effect, 11 following the recommendations of the Advisory Committee that commenced work on the issue in 1999. 12

The most significant change to the Federal Rules brought about by the amendments was the establishment of a framework for parties to give early attention to the question of e-discovery. 13 Parties are required to confer early on in the proceedings about ESI, in order to resolve issues such as the form in which the information will be produced, issues regarding the retention of ESI and steps to deal with inadvertent waivers of privilege. 14

The amendments also confirm (for the avoidance of doubt) that the obligation of parties to disclose information includes potentially relevant ESI in their custody. 15 At the same time, the amended rules allow for a proportionality test for the production of ESI, stating that ESI need not be produced if it is "not reasonably accessible because of undue burden or cost" (the burden of proving inaccessibility resting on the producing party), unless the requesting party shows "good cause" to order such discovery. 16

With regard to requests to produce, the amended rules make provision for the requesting party to specify the form in which they wish ESI to be produced (subject to a right of the producing party to object) and to ask to "inspect, copy, test or sample" ESI belonging to the opposing party. 17 Finally, [Page379:] the amendments contain a so-called "safe harbour" provision, i.e. an exception to the penalties for spoliation of data, by providing that, absent exceptional circumstances, a party may not be penalized for loss of data as a result of the routine, good-faith operation of an electronic information system. 18

At around the same time, changes were also adopted to the practice direction of the Civil Procedure Rules in England and Wales ("CPR"). A working party chaired by Cresswell J. was set up under the auspices of the Commercial Court Users' Committee, which produced a report dated October 6, 2004. 19 This led to the adoption of a new Section 2A of the Practice Direction to Part 31 of the CPR on October 1, 2005. 20

The English amendments were more modest in nature than the amendments to the Federal Rules, reflecting the fact that the scope of disclosure required to be given by parties before the English courts is narrower than the discovery required of American parties. In addition, the English courts had already gone a considerable way towards addressing the issue of disclosure of ESI in their case law, recognising that ESI falls within the class of "documents" that are disclosable and adopting a hands-on case-management approach which ensured that ESI was treated proportionately.

The amendments confirmed, for the avoidance of doubt, that the definition of a "document" in the CPR includes ESI, even documents that are stored on servers and back-up systems and electronic documents that have been "deleted". 21 As with the Federal Rules, the English amendments made provision for the parties to confer early on to discuss disclosure and preservation of ESI, including agreeing on what format ESI should be in when produced. 22 Finally, the amendments note that ESI can have an effect on the balancing test carried out by the parties and the judge when deciding on the extent of a "reasonable search" for discoverable material and seek to set out guidance for how this impact is to be managed. 23

The concern to address the issue of ESI has also been felt in other common law jurisdictions. For example, the Federal Court of Australia has introduced e-discovery and e-filing reforms.

Interestingly, many of these changes in common law legislative systems were motivated by the aim of limiting the damaging effect, in terms of time and cost, that ESI discovery or disclosure could have on the litigation process if specific guidelines were not enacted. [Page380:]

3. The evolution of arbitration-specific guidelines

The vigorous debate that has surrounded the amendments to various court procedural rules has inevitably spilled over into the world of arbitration. As has been forcefully pointed out by arbitration practitioners of all backgrounds, 24 the questions that arise in an arbitration context are not the same as those in a litigation context.

Arbitration is not litigation-and it has an interest in preserving its character as an alternative forum that can be more responsive to the needs and expectations of its users than a court system.

As a matter of fact, the most celebrated US case on e-discovery, Zubulake, 25 involved a claim by an employee for gender discrimination and illegal retaliation over which arbitrators would normally have no jurisdiction. It will often be the case in such a claim that one party (the employer) holds far more by way of documentation and information than the other party (the employee). In international commercial arbitration, where the parties are often both businesses, there is frequently (although by no means invariably) a greater equality of arms in terms of documentation held. In a contractual dispute, for example, it is likely that both sides, if they have been conscientious in their record keeping, will have a copy of the contract, all the drafts of it that have been exchanged and all the correspondence between them about the contract. It stands to reason, therefore, that the "best practice" for production of documents in arbitration differs from that developed in the litigation context.

While practices concerning document production in international arbitration can vary widely, it is fair to say that orders for general discovery or standard disclosure are the exception to the rule. It is usual, in international arbitration, for orders for production of particular documents or a class of documents to be made at the request of one of the parties, after the party has satisfied the tribunal of the relevance of the documents in question to the outcome of the case and therefore after the first exchange of briefs. As a result, the way in which court procedure in various common law jurisdictions has tackled the question of ESI does not represent the right solution in the international arbitration context.

Nevertheless, some commentators have called for the adoption of detailed guidelines on the production of electronic documents in arbitration, 26 and various arbitral bodies and other organizations are discussing the usefulness of enacting such guidelines. Recent developments include: [Page381:]

Discussion in the legal press about the need for amendment of the IBA Rules, 27 and a survey currently under way by the IBA Rules of Evidence Subcommittee to determine whether there is a need for changes to the rules, including in relation to ESI.

A Task Force on Production of Electronic Documents in Arbitration set up by the ICC in August 2008, mandated to study the disclosure of electronic documents in international arbitration and establish a report, possibly in the form of notes or recommendations. 28

Guidelines for Arbitrators Concerning Exchanges of Information, issued by the International Centre for Dispute Resolution (ICDR), effective May 31, 2008, which include a short section on electronic information. 29

A Protocol for E-Disclosure in Arbitration, which was issued by the Chartered Institute of Arbitrators (CIArb) on October 2, 2008. 30

The International Institute for Conflict Prevention and Resolution is working on a draft protocol on the production of information, including ESI.

The very title of the CIArb Protocol, "E-disclosure in Arbitration", and to a certain extent its content, suggests that disclosure, which has a specific legal meaning under English law, is a feature of international arbitration. As already stated, the author does not share that analysis.

However, and whatever one may think of the CIArb Protocol on E-disclosure, these guidelines and others can serve to draw the attention of parties and tribunals to the issues involved in the production of ESI and can suggest solutions that may be adopted (when appropriate) to deal with these issues.

Nevertheless, the specificity of the issues arising in each particular case means that guidelines should not detract from the flexibility of counsel and tribunals to deal with the issues that arise on a case-by-case basis. The responsibility for sensible case management rests with the parties and tribunals themselves. [Page382:]

4. Dealing with ESI in an arbitration context

With this in mind, the following examination sets out four principles that parties and tribunals should bear in mind when faced with issues relating to the production of ESI:

Production of ESI should be kept in perspective-it is at base just another form of document production and should be dealt with in the way that best serves the parties' interests.

ESI should encourage parties and tribunals towards something that is in any event a good idea: pro-active case management.

Given the need for the tribunal to ensure that parties' expectations are met and fairness and proportionality are maintained, the issue of ESI may call for tailored solutions to be adopted on a case-by-case basis.

There are some specific pitfalls that are more likely to arise when dealing with ESI, which parties and tribunals should bear in mind.

The present author is of the view that, assuming it is properly handled, electronic evidence should not "change the face of arbitration". Likewise, ESI should not be an argument in favour of introducing discovery, in the American sense of the word, as a feature of international arbitration-on the contrary, it should have a deterrent effect.

The following analysis is intended to contribute to the debate on best practice relating to ESI, the outcome of which will hopefully be to find ways of preserving the role of arbitration as a user-friendly and efficient alternative to court litigation well into the 21st century and beyond.

a. Keeping ESI in perspective

So much has been said and written about the problems relating to the exchange of electronic evidence that the question of how to manage ESI has taken on a life of its own.

When faced with the issue in practice, tribunals and parties would therefore do well to return to first principles and remind themselves that the object of document production is to obtain evidence that will be useful to the tribunal in determining the questions before it, while at the same time [Page383:] meeting party expectations as to how much time and money they want to spend on the proceedings. Despite its unique characteristics, as discussed above, electronic evidence is simply another form of evidence, which can be used by a party to help it to prove its case.

The usual practice for arbitrators faced with a request to produce is to consider the relevance of the requested material and the proportionality of the request-an approach that is reflected in the IBA Rules. These concepts are of course material to all requests for production of documents, but they become particularly significant in the context of ESI.

The ease with which electronic information can be created, stored and disseminated means that, invariably, large quantities of ESI held by a party will be of no practical use to the case that either party has to prove or to the tribunal in deciding the issues before it. For example, a request for production of e-mails could turn up voluminous amounts of irrelevant correspondence (including "conversation" which in the past would probably have taken place orally). Likewise, a keyword search may turn up large numbers of documents that respond to the request by coincidence but are irrelevant to the dispute.

It is therefore important for parties, when making a request for production of ESI, and tribunals, when deciding such a request, to reflect particularly carefully about the issues at stake in the case and how the information requested can contribute to deciding them. At the same time, parties and tribunals should take particular care in this context to look at a request for production of documents in the light of the amount at stake in the case, in order to ensure that the costs of production do not engulf the claim.

The goal should be to ensure that electronic evidence is treated proportionately, with orders for production being targeted at what is really necessary to help the tribunal resolve the dispute before it in a fair manner. As a result, ESI should reinforce both the test of relevance of the requested material and that of proportionality. ESI should be neither a strategic tool that one party wields against the other to instil fear, increase its costs and bully it into settling, nor an unwelcome surprise that derails the procedural timetable because the parties had not thought to deal with it in good time. [Page384:]

b. Pro-active case management

As recognized in the Federal Rules, the CPR and the arbitration guidance on the subject, the key to managing ESI in cases where it is likely to be an issue is awareness and early intervention. Dealing with searches and review of ESI where it is voluminous can be extremely time-consuming. If these processes are not set in train early enough, they can easily disrupt the "critical path" of hearing preparation and lead to unnecessary postponements to the hearing date.

If ESI appears to be an issue in a particular case, it should be discussed between the parties before the first procedural hearing, and appropriate orders should be applied for and made promptly if an agreement between the parties has not been possible. The later the issue is raised, the greater the potential for serious interference with the procedural timetable.

The nature of ESI is such that even when agreement seems to have been reached by parties at an early stage, unresolved issues may later arise and pose problems. For example, production of documents may uncover a pressing need for further information-a seemingly fraudulent document may lead a party to make a request for production of metadata or a request for documents in a different format from that originally produced. Likewise, with the best intentions, parties may not have a full understanding of the importance of ESI to their case at the time of the first procedural hearing, so later orders may be required.

Pro-active case management is therefore an ongoing process, and parties need to remain aware of the need to address questions relating to ESI throughout the case.

c. Tailored solutions

Coping with electronic evidence has become a veritable industry in the United States and, to a lesser extent, in other common law jurisdictions. This leads to the ongoing development of new technological solutions, as well as innovative procedural orders to deal with issues particular to ESI. An awareness of these developments as they occur can only be of assistance in the world of arbitration. Arbitrators and in-house and external counsel do not need to be IT specialists, even less specialists on US litigation, but they do need to develop an understanding of the problems that can arise when a party holds large amounts of ESI and the tools that can be used to tackle them. [Page385:]

In any event, awareness of these technicalities will primarily be useful in the process of gathering a party's own evidentiary material.

i. New technologies

Search and retrieval software has undergone rapid development and improvement in recent years, impelled partly by internet search engines such as Google, Yahoo and eBay. The e-discovery support industry has made use of this growth area to develop new software particularly targeted to the legal context.

There are a range of products on the market that allow parties to manage, group and review large databases of documentation held in different formats. The ability to find relevant information through search software can be particularly useful in reducing reviewer costs. A variety of options exist:

The simplest is a keyword search-the parties to a dispute agree (or the tribunal decides) on a list of keywords as well as the extent of the database or databases to be searched, and the resulting documents (once they have been filtered for privileged and any other undisclosable information) are produced to the other side.

A slightly more complex variant of this is the use of a search engine that supports "Boolean" logic to combine keywords and "operators" (such as "AND", "OR" and "AND NOT" or "BUT NOT"), which allows for the search of combinations of keywords.

There is also a growing number of more sophisticated tools in the market place, which use mathematical algorithms, statistics and various linguistic techniques to help find, group and present related content. For example, the use of "fuzzy logic" techniques and "conceptual search methods" can help to ensure that the keyword search better reflects the ways in which language is actually used by human beings (searching for the plural as well as the singular, verbs as well as nouns, erroneous spelling, words that are semantically connected and so on).

In addition, "de-duplication" software can narrow the search area by removing the redundant duplicate copies of documents or e-mails from the search pool. [Page386:]

The limitations of these programmes should be borne in mind. The use of search engines and techniques is necessarily inaccurate and can lead to the production of irrelevant information or equally to the omission of documents that are highly relevant. While some of the new software referred to above can assist in improving accuracy, it is inevitably difficult to tell whether the algorithms used are truly useful in the circumstances of a particular case. These risks will mean that keyword searches will only be useful in certain situations-for example, where the information concerned is highly technical or where the database involved is too large for any other form of search to be feasible.

For arbitration counsel, software tools will initially be of use in examining the documentation held by their client for material to support their own case. Their usefulness in dealing with requests to produce will in many cases be secondary, but in other cases software tools can be the only way of responding quickly and efficiently to a request to produce.

ii. Innovative procedural orders

Secondly, there are solutions to the potential difficulties caused by ESI that have proved useful either in the context of litigation or in arbitrations that have had to deal with the issue.

Counsel and arbitrators who keep their fingers on the pulse of such developments will have a useful resource to draw from when the need arises. As already mentioned, there are significant differences between arbitration and litigation. This means that solutions developed to serve the needs of a common law court procedure cannot simply be adopted into the arbitration context wholesale. At the same time, an informed understanding of these possibilities can assist tribunals and parties in dealing with new issues as and when they arise.

A few examples of procedural orders tailored to the production of ESI that have been used by the courts and can be adopted (with appropriate modifications) in the arbitration context are described below:

Orders can be made specifying that the production of documents should take place in a particular form. While production in hard copy is still the norm in arbitration, this can be unwieldy and will deprive the receiving party of the opportunity to have a searchable database with metadata information intact. There are therefore circumstances where other forms of production are appropriate. For example, production in native format can allow access to metadata where there is a particular need for this to be seen. It can also be [Page387:] the best way of properly understanding a document-for example, a spreadsheet is best understood in its native format. Similarly, an order for production in TIFF or PDF can allow the parties to carry out bates numbering (i.e. sequential numbering of each page with a unique number).

The parties or tribunal may choose to identify, early on in the proceedings, certain individuals who are "key players" for the purposes of the dispute. This can assist reflection on orders for production of documents, searches, as well as any steps that may need to be taken to ensure the preservation of documents. An order for preservation or disclosure of documents generated by a "key player" will often not be appropriate in an arbitration context. An arbitrator will often wish to be satisfied of the relevance of the documents in question before making an order to disclose-and it is virtually certain that not all of the documents generated by the "key player" will be relevant to the case. However, this solution may be useful in certain types of arbitration, in particular where both parties have an expectation that there will be comparatively extensive document production.

Where the parties are faced with voluminous amounts of ESI and are not certain of the relevance or value of documents contained in the database, orders for "sampling" of carefully chosen representative sections of the database can help parties to understand the nature of the database and allow them to make a subsequent focused request for production. An order for production of "sample" documents will once again not be granted by many arbitrators, because the requesting party is (by definition) not sure of the relevance of the documents in the "sample". An arbitrator may therefore dismiss the request because he is not satisfied that the documents should be produced on the grounds of relevance. Once again, however, this is a solution that will be useful in certain types of case and given certain party expectations.

Lastly, there are circumstances where an arbitrator may wish to take production of documents issues into account when awarding costs.

d. Awareness of the pitfalls

Finally, both the volume in which electronic information is likely to be held and the particularities of the way in which it is stored can give rise to particular issues that should be considered when a request for production of electronic evidence is made. Two examples of such issues are the potential for inadvertent production of privileged or other undisclosable material and the potential for breach of data protection. [Page388:]

i. Material that is inadvertently disclosed

Experience in the context of US litigation has shown that inadvertent disclosure of privileged material is more likely to happen where the parties hold their documents electronically. In part, this is simply because of the volume of material disclosed, but the problem can also be exacerbated by the use of tools such as keyword searches.

Much attention in the US context has focused on the question whether, if privileged documentation is inadvertently disclosed in the course of e-discovery, this should be treated as a waiver of privilege, either for the individual document or for related documents on the same issue. In general, when deciding whether a waiver of privilege has occurred, a US or English court will examine the degree to which the privileged material can be said to have been disclosed intentionally and the extent to which the disclosing party took reasonable precautions to avoid disclosing materials in error.

The same issue of inadvertent disclosure of privileged material or documents subject to a civil law secret professionnel can arise in arbitration, albeit that the status of these documents in an international arbitration is complicated by the fact that different legal regimes offer different types of protection to documents generated during the legal process. 31 The task of an arbitrator faced with a claim for privilege or secret professionnel is therefore to attempt to do justice to the expectations of the parties and their counsel, given their different legal backgrounds.

The issue of inadvertent disclosure in international arbitration may also arise in other contexts-for example, a party may inadvertently disclose documents that it would have wished to claim were confidential, trade secrets or politically or institutionally sensitive. As a matter of general practice (which is reflected in the IBA Rules), a tribunal will take matters such as these into consideration when deciding whether or not to order production of documents. If these documents are inadvertently disclosed, the tribunal will have to decide whether or not it should allow the receiving party to rely on the information. The considerations for making this decision are similar to those applied in the context of privilege. Where the question arises in the context of the production of ESI, the tribunal will need to take into account the context of the disclosure in conducting its balancing exercise (Was the document disclosed as part of a voluminous batch of information that made failsafe checking impossible? Was it disclosed in response to a keyword search?). [Page389:]

ii. Data protection and privacy

A factor that is often overlooked when, in the thick of a case, parties transfer documents to each other is the potential for breach of data protection and privacy legislation. 32 Once again, the possibility of this occurring increases exponentially where large amounts of data are transferred between the parties, as can be the case when ESI is produced.

In Europe, the 1995 EU Data Protection Directive33 protects personal data, which is broadly defined to include any information relating to an identifiable individual. Ascertaining the precise nature of the protection covering particular data can be troublesome, because it can be difficult to decide in which jurisdiction a particular piece of electronic information is held for data protection purposes, and because the level of protection granted inside and outside the European Union can be very different. 34 For example, the United States has no comprehensive data protection legislation, relying instead on a sectoral approach, with a mix of legislation, regulation and self-regulation.

The starting point regarding the transfer of personal data across borders under the directive is that it is generally allowed within the European Economic Area, but disallowed where the transfer is to a receiving country that provides inadequate levels of protection for such data. There are certain limited derogations from this principle.

Firstly, a transfer of data to an "unsafe" state is permitted (subject to any national rules applying stricter criteria) where the transfer is "necessary or legally required on important public interest grounds, or for the establishment, exercise or defence of legal claims". 35 The standard of "necessity" is a high one-it requires, for example, that adequate steps are taken to ensure that only relevant materials are transferred (not, for example, irrelevant private correspondence that happens to have been picked out by a key word search).

Secondly, a transfer of data to an "unsafe" state may also be possible where the data controller uses other means to ensure that the data will be adequately protected after transfer, for example by entering into contractual arrangements with the party receiving the data. The European Commission has published certain standard-form contractual clauses that are deemed to meet these requirements. [Page390:]

The relevant domestic laws may also provide additional protection against transfers of data that breach an employee's right to private life. For example, in France, an employer is not entitled to access personal (i.e. non-professional) e-mails sent by an employee unless it has obtained his or her consent or an order of the court allowing it to do so on good grounds. This is the case even if the e-mails are sent from a work computer and the employer had prohibited the employee from using the computer for non-professional purposes. 36 Case law has not provided a clear explanation of the distinction between professional and personal e-mails, which is determined on a case-by-case basis. This places real difficulties in the way of a company that, in an arbitration, produces materials that may include personal ESI (for example, an e-mail box that may contain personal e-mails), where the company is unable to obtain consent from the relevant employee or former employee.

As far as ESI production in an international arbitration process is concerned, the best solution is invariably to ensure that, where possible and particularly where it is not likely to prove relevant to the issues in the case, personal data is excluded from any document production.

5. Conclusion

It is fair to say that electronic evidence has changed the face of court procedure, at least in the United States, England and other common law jurisdictions. Judges are compelled to pay more attention to the proportionality of standard disclosure and requests for production of a particular class of documents. The new court procedural rules mean that, in practice, courts are required to be more receptive to arguments about whether discovery or disclosure is an undue burden on the producing party or whether the documents requested are truly relevant to the issues at hand.

As a result, court disclosure orders are more carefully tailored-and even, one might say, more similar to the type of order to produce documents that one might traditionally have encountered in an international arbitration.

It is to be expected that electronic evidence will also have an influence on the way arbitrators and parties think about the production of documents as well as on the examination by each party of the evidentiary material it has under its control. But will e-discovery (understood as US-style discovery of electronic documents) become a feature of international arbitration? When the parties want it, it will, but in the absence of this, the present author thinks not. The existence of electronic evidence is very far from being a [Page391:] reason for arbitration to move towards a fuller discovery process. If anything, the excesses encountered in US litigation in the field of e-discovery give arbitration practitioners more of an incentive to keep the issue in perspective.

Indeed, it would be more than a paradox if ESI, which has prompted a relative limitation of the scope of discovery in common law courts, had the reverse effect in international arbitration and enhanced the recourse to discovery in cases where it is currently virtually unknown.

And will electronic evidence change arbitration beyond all recognition? Again, the present author thinks not. Electronic evidence puts a range of new tools at the disposal of parties and arbitrators. Used properly, they will simply be other means of doing what arbitration has always striven to do, namely to resolve disputes between parties in a fair, efficient and cost-effective way.

1
I refer, for example, to the recent Juris Conferences in New York and London, which dedicated a full day seminar to the question of "Electronic Evidence and Disclosure in International Arbitration" and led to the publication of a book entitled Electronic Disclosure in International Arbitration (Jurisnet LLC 2008).

2
I note at the outset that the words "discovery" and "disclosure" have specific technical meanings in the US and English rules of court. In the United States, "discovery" refers to the right of a party to obtain all documents that are "relevant to the claim or defense of any party". It can even include inadmissible evidence where such evidence is "reasonably calculated to lead to the discovery of admissible evidence". "Discovery" in the United States also refers to the right to obtain oral depositions of fact and expert witness, written interrogatories and inspection of property. In England, the term "discovery" was previously used, but it has now been replaced with the term "disclosure". The new procedural regime brought in by the Civil Procedure Rules 1998 gives parties a quasi-automatic right to disclosure of all documents in the possession, custody or control of another party on which that party relies, or which adversely affect that party's case or support another party's case. On top of this so-called "standard disclosure", parties can also apply for the production of specific documents or classes of documents. Neither "discovery" nor "disclosure" in the sense referred to in the rules of court is normally ordered in international arbitration. In an arbitration context, it is therefore clearer to refer to "exchange of evidence" or "document production" to ensure that the terms used are neutral.

3
MacKinnon v. Donaldson Lufkin & Jenrette Securities Corp and others [1986] 1 All ER 653, per Hoffman J.

4
For an instructive exploration of the differences between the civil law (in particular French) and common law approaches to taking evidence from the perspective of a common lawyer, see James Beardsley, 'Proof of Fact in French Civil Procedure', 34 American Journal of Comparative Law (1986) pp. 459-486. For the civil law perspective, see Matthieu De Boisseson, 'Taking of Evidence in the Arbitral Proceedings: Comparative Introduction to the Systems of Producing Evidence in Common Law Countries and Countries of Roman Law Tradition', Dossier of the Institute of International Business Law and Practice, No. 8/1989, pp. 101-111.

5
See, in particular, Bernard Hanotiau, 'Document Production in International Arbitration: A Tentative Definition of Best Practices', Document Production in International Arbitration, ICC International Court of Arbitration (2006) Special Supplement.

6
Article 3 of the IBA Rules on the Taking of Evidence in International Commercial Arbitration.

7
See, for example, the 1974 decision of the English High Court, Grant v. Southwestern and County Properties Ltd [1975] Ch 185 at 197, where Walton J. held that: "The mere interposition of necessity of an instrument for deciphering the information cannot make any difference in principle. A litigant who keeps all his documents in microdot form could not avoid discovery because in order to read the information extremely powerful microscopes or other sophisticated instruments would be required. Nor again, if he kept them by means of microfilm which could not be read without the aid of a projector."

8
For an in-depth exploration of the features of ESI that can create difficulties for international arbitrators, see John M. Barkett, 'E-Discovery for Arbitrators', 1(2) Dispute Resolution International (2007) pp. 129-170.

9
'The Sedona Principles: Best Practices Recommendations & Principles for Addressing Electronic Document Production', first published in January 2004, second edition June 2007, available at: http://www.thesedonaconference.org/content/miscFiles/ TSC_PRINCP_2nd_ed_607.pdf.

10
'The Sedona Conference Best Practices Commentary on the Use of Search & Information Retrieval Methods in E-Discovery', 8 The Sedona Conference Journal (2007). The Sedona Conference based its calculation on the assumption that one gigabyte equals 80 000 pages and that an associate billing US$200 per hour can review 50 documents per hour at 10 pages in length. This figure should be seen in the context of the fact that a single storage device (e.g. a personal computer) may hold many gigabytes of information and the server or set of servers for a large organization would typically have a capacity of several thousands of gigabytes.

11
The amendments to the Federal Rules are accessible on the US courts website at: http://www.uscourts.gov/rules/supct1105/CV_Clean.pdf.

12
The relevant excerpts from the Advisory Committee Report are available at: http://www.uscourts.gov/rules/supct1105/Excerpt_CV_Report.pdf.

13
See the Federal Rules, Rules 16, 26(a), 26(f) and Form 35.

14
Ibid., at Rule 26(f).

15
Ibid., at Rule 26(a).

16
Ibid., at Rule 26(b)(2).

17
Ibid., at Rule 34(a) and (b).

18
Ibid., at Rule 37(f).

19
The Cresswell Report is available on the Commercial Court website at: http://212.137.36.113/publications/misc/admiralcomm/index.htm.

20
CPR and Practice Directions can be found on the website of the Department of Constitutional Affairs. for the Practice Direction to Part 31, see: http://www.dca.gov.uk/civil/procrules_fin/contents/ practice_directions/pd_part31.htm.

21
Practice Direction, para. 2A.1.

22
Ibid., at paras. 2A.2 and 2A.3.

23
Ibid., at paras. 2A.4 and 2A.5.

24
Perhaps most eloquently by Michael E. Schneider in his recent article, 'A Civil Law Perspective: "Forget E-Discovery"', Electronic Disclosure in International Arbitration (Jurisnet LLC 2008).

25
Zubulake v. UBS Warburg, 217 F.R.D. 309 (S.D.N.Y. 2003).

26
See, for example, Robert Smit and Tyler Robinson, 'E-Disclosure in International Arbitration', 24(1) Arbitration International (2008) at p. 120.

27
See, for example, Nicholas Tse and Natasha Peter, 'Confronting the Matrix: Do the IBA Rules Require Amendment to Deal with the Challenges Posed by Electronically Stored Information?', 74(1) Arbitration (2008) at p. 29.

28
See: http://www.iccwbo.org/policy/arbitration/index.html?id=23620.

29
Available at: http://www.adr.org/si.asp?id=5288.

30
Available at: http://www.arbitrators.org/institute/CIArb_e-protocol_b.pdf.

31
See Irene C. Warshauer, 'Electronic Discovery in Arbitration: Privilege Issues and Spoliation of Evidence', 61(4) Dispute Resolution Journal (November 2006-January 2007).

32
The potential conflicts surrounding data protection that can arise in the litigation context have been addressed in the recent publication by the Sedona Conference, 'Framework for Analysis of Cross-Border Discovery Conflicts: A Practical Guide to Navigating the Competing Currents of International Data Privacy & E-Discovery-Public Comment Version', published in August 2008 and shortly to be accompanied by a companion publication: 'The Sedona Overview of International E-Discovery, Data Privacy and Disclosure Requirements'.

33
Directive 95/46/EC of the European Parliament and of the Council on the Protection of Individuals with Regard to the Processing of Personal Data and on the Free Movement of Such Data.

34
In France, the Commission Nationale de l'Informatique et des Libertés (CNIL) issued a press release on 18 January 2008 noting a widespread concern about the interaction between the discovery process and the protection of personal data (available at: http://www.cnil.fr/index.php?id=2379&news[uid]=512&cHash=0b7756c4aa). The CNIL has established a working group that is in the process of hearing representations from interested parties. The subject is also under discussion at the European level in the Article 29 Data Protection Working Party.

35
See Article 26(d) of the Directive.

36
See, for example, Law No. 91-646, 10 July 1991, Journal Officiel No. 162, 13 July 1991; Cass. soc., 2 October 2001, No. 99-42.942, SA Nikon France v. Onof, Bull. civ. V, No. 291.