1. INTRODUCTION1

International arbitration cannot withstand the inexorable march of technological progress any more than the national courts with which it competes for jurisdiction.2 As recently as this month, Clifford Chance – a leading law firm – announced a comprehensive training programme for its lawyers that covers cybersecurity, block chain technology and predictive coding.3 In this article, I will argue that predictive coding can, and should, bring efficiency gains and costs savings to the document production phase, and thus the arbitration process overall.4 given the flexibility already inherent within the arbitral process, predictive coding can, and should, fit into proceedings without any need for changes to arbitration laws or institutional rules.5 In all likelihood, it is already doing so. examples of quick adoption in other jurisdictions indicate that predictive coding is likely to become commonplace sooner than many practitioners expect.

2. WHAT IS PREDICTIVE CODING?

2.1 Summary

Predictive coding is a software algorithm that can be 'taught' how to interrogate a set of electronic data in order to identify documents possessing certain characteristics with a high degree of accuracy. There are three main ways in which it can be used:

  1. It can replace extensive human review. Initially, a senior lawyer with proper knowledge of the issues in a given case reviews a stratified sample of documents and codes them. The software 'learns' from this coding, then replicates that coding across most of the remaining data set.
  2. It can make human review more efficient. As all document reviewers progress through search results, the software continually 'learns' from their cumulative coding and 'promotes' un-reviewed documents to ensure that the next documents to be reviewed by humans are more likely to be relevant.
  3. It can 'quality control' human review. As all document reviewers progress through search results, the software 'learns' from their actual coding but simultaneously in the background records the coding it thinks should be applied in light of the actual coding cumulatively applied so far. At the end of the review, the two sets of results can be juxtaposed and the differences between them reviewed to ensure nothing has been missed or coded in error.

while useful, the second and third ways are a matter of internal case management, which need not involve the other party to proceedings. Although both yield efficiency and risk mitigation benefits, they still involve extensive human review and so do not bring the huge advantages described below. This article therefore focuses only on the first way, which, as studies show,6 can produce superior, more consistent results than manual document review.

2.2 Commentary

Perhaps unsurprisingly given the relative novelty of predictive coding, there are no publicly available awards that mention it, though it may already have featured in some procedural orders.7 In addition, predictive coding is not yet mentioned in the document production chapters of several leading practitioner texts. given its swift adoption in several jurisdictions, however (on which see below), it soon should be.

In the meantime, these same practitioner texts all emphasise:

  1. The unwieldy nature of the vast numbers of electronic documents that often need to be searched during disputes;8
  2. The different expectations in respect of the approach to evidence gathering of arbitration practitioners from different jurisdictions;9
  3. The flexibility available to parties in their approach to evidence and document production;10 and
  4. The extensive discretion tribunals possess in this regard.11

These factors, on which there is academic (and practitioner) agreement, suggest that widespread usage of predictive coding should be a possible and welcome development in international arbitration.

Footnotes

1 This material is the author's own and does not purport to represent the positions, strategies or opinions of Milbank Tweed hadley & McCloy LLP.

2 A good introductory summary of technologies that could become more prevalent in disputes is: Lucas Bento, 'International Arbitration and Artificial Intelligence: Time to Tango?', Kluwer Arbitration Blog, february 23 2018.

3 M. Munro, 'Clifford Chance offers coding training to all lawyers', The Lawyer, 5 April 2018.

4 See e.g. B. Krużewski and R. Moj, 'Documents in Construction Disputes', The Guide to Construction Arbitration (eds. S. Brekoulakis D. Brynmor Thomas), Global Arbitration Review, 2017

5 See e.g. D. Kiefer and A. Cole, 'Suitability of Arbitration Rules for Construction Disputes', The Guide to Construction Arbitration (eds. S. Brekoulakis D. Brynmor Thomas), Global Arbitration Review, 2017

6 See e.g. M. grossman and g. Cormack, 'Technology assisted review in e-discovery can be more effective and more efficient than exhaustive manual review', Richmond Journal of Law and Technology, Volume 17, Issue 3, 2011, 1-48.

7 This is to the best of my knowledge, but if any readers know differently it would be interesting to hear from you! For an understanding on how predictive coding has been applied in english litigation, see the recent decision in Pyrrho Investments Limited & Anr v MWB Property Limited and Others [2016] ewhC 256 (Ch). This represents the first time an english court has provided a detailed judgment closely examining and approving the use of predictive coding, as explained further on in this article.

8 N. Blackaby and C. Partasides, Redfern and Hunter on International Arbitration, oxford University Press, 2015 ("Redfern and Hunter"): 6.104-110; D. Sutton, J. gill and M. gearing, Russell on Arbitration, Sweet & Maxwell, 2015 ("Russell"): 5.108, 5.142-144; g. Born, International Commercial Arbitration, wolters Kluwer, 2014 ("Born"): 16.02[e][5].

9 Redfern and hunter at 6.77-79, 6.92-94; Russell at 5.109-110; Born at 16.02[e][1].

10 Redfern and hunter at 6.01-06; Russell at 5.100, 5.144-145, 5.147; Born at 16.02.

11 Redfern and hunter at 3.229; Russell at 5.105, 5.144-145, 5.147; Born at 16.02.

Please click here to read the full report.

Previously published in Young Arbitration Review.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.