Dispute Resolution Beijing/Hong Kong/Shanghai Client Alert Use of Predictive Coding in Electronic Discovery in England Provides Useful Reminder for Hong Kong Cases Recent Developments In a recent landmark decision concerning the discovery of electronic documents (“English Case”), the English High Court approved the use of a predictive coding program to narrow the scope of relevant documents. The development is significant for litigants in Hong Kong, as this is the first time an English court has approved the use of predictive coding in the discovery of electronic documents (“e-discovery”). Predictive coding is an electronic document review process involving the use of a machine learning algorithm to distinguish relevant from nonrelevant documents, based on a human reviewer’s coding of sample sets of documents. The English Case provides a useful reminder for the potential use of predictive coding in Hong Kong court proceedings. The Hong Kong’s Practice Direction “Pilot Scheme for Discovery and Provision of Electronically Stored Documents in Cases in the Commercial List” (“PD SL1.2”), which was modelled on its 2010 UK equivalent, explicitly refers to the use of technology assisted review (“TAR”). Implications for Hong Kong Because of the complex nature and volume of electronic documents, e-discovery has resulted in the development of specialised tools to filter relevant documents, including predictive coding and other forms of TAR. The English Case confirms that the use of predictive coding programs can lead to more proportionate and accurate disclosure. To recap, the main objective of PD SL1.2 is to ensure proportionality and cost-effectiveness. Litigants are required to conduct a “reasonable search” of electronic documents for the purpose of e-discovery. What is reasonable depends on various factors, including the nature of the proceedings, the number of electronic documents and the ease and expense of their retrieval, and the significance of any electronic documents likely to be located during the search. In addition, PD SL1.2 also directs the parties to discuss the use of tools and techniques to reduce the burden and cost of e-discovery. Although PD SL1.2 applies only to cases in the Commercial List and cases where the parties agree to follow it or where the court so directs, the Hong Kong court has indicated that it may apply the general principles of PD SL1.2 to e-discovery in cases to which PD SL1.2 does not apply (Chinacast Education Corp v Chan Tze Ngon  HKEC 1381). The English Case The English Case concerned a shareholder dispute involving claims in the tens of millions of pounds. Most of the relevant documents were contained April 2016 www.bakermckenzie.com Beijing Suite 3401, China World Office 2 China World Trade Centre 1 Jianguomenwai Dajie Beijing 100004, PRC Tel: +86 10 6535 3800 Fax: +86 10 6505 2309 Hong Kong 14th Floor, Hutchison House 10 Harcourt Road Central, Hong Kong Tel: +852 2846 1888 Fax: +852 2845 0476 Shanghai Unit 1601, Jin Mao Tower 88 Century Avenue, Pudong Shanghai 200121, PRC Tel: +86 21 6105 8558 Fax: +86 21 5047 0020 2 Baker & McKenzie | April 2016 in email accounts stored on back-up tapes and totalled around 3.1 million after duplicate files were removed (a process called “de-duplication”). The parties sought court approval to use a predictive coding program with the following workflow: • The parties set criteria (e.g. date range, custodians) for inclusion of documents in the review. A human reviewer reviews a representative sample of documents, coding them as relevant or not relevant. The computer program analyses the sample documents and codes them as relevant or not relevant. The reviewer then reviews the accuracy of the program’s coding. • Documents correctly coded as relevant are further analysed by the program, whilst documents incorrectly coded as relevant are removed from the sample. • The parties agree in advance the data set, sampling size, batches, reviewer, confidence level and margin of error. The sampling process is repeated as many times as necessary to bring the program’s correct coding decisions to an agreed percentage. • Once the program is trained, it applies the sampling results of the sampling across all documents, and scores and prioritises each document as to its likely relevance to the issues in the case. The English court obtained guidance from US and Irish decisions, and approved the use of predictive coding technology based on factors including: • Evidence that the use of predictive coding programs leads to more accurate disclosure rather than manual review alone or keyword searches and manual review combined; and • The estimated costs of using a predictive coding program were proportionate to the value of claims, as opposed to a full manual review. Actions to Consider While e-discovery tools are getting more sophisticated and enabling more intelligent searching, the English Case is a timely reminder for companies to review their policies regarding electronic data (including their use, processing and storage). A few practical steps which companies can take include: • Review document management and retention policies, as well as policies for collection of data from employees, including data held on company issued computers and mobile devices; • Take a data inventory and identify what data you have, where it is stored and who has access to it; and • Implement a litigation plan, including protocols for the preservation and collection of electronic documents when faced with potential court proceedings, and training for IT staff when faced with data retention and collection. Conclusion The discovery process in Hong Kong legal proceedings is still guided by Order 24 of the Rules of High Court. E-discovery only applies to situations specified in PD SL1.2 and mainly cases in the Commercial List, but its application will undoubtedly expand given the developments in other jurisdictions. Although the English Case is not binding upon Hong Kong courts, the Hong Kong courts will likely adopt a similar approach to the English court, and allow predictive coding to be used in e-discovery in appropriate cases. Should you wish to obtain further information or want to discuss any issues raised in this alert with us, please contact: Cynthia Tang Partner +852 2846 1708 firstname.lastname@example.org Bryan Ng Partner +852 2846 2923 email@example.com Philipp Hanusch Associate +852 2846 1665 Philipp.Hanusch@bakermckenzie.com Michelle Tennant Associate +852 2846 2576 firstname.lastname@example.org This publication has been prepared for clients and professional associates of Baker & McKenzie. Whilst every effort has been made to ensure accuracy, this publication is not an exhaustive analysis of the area of law discussed. Baker & McKenzie cannot accept responsibility for any loss incurred by any person acting or refraining from action as a result of the material in this publication. If you require any advice concerning individual problems or other expert assistance, we recommend that you consult a competent professional adviser. Unsubscribe To unsubscribe from our mailing list or to change your communication preferences, please contact email@example.com. © 2016 Baker & McKenzie. All rights reserved. Baker & McKenzie International is a Swiss Verein with member law firms around the world. In accordance with the common terminology used in professional service organizations, reference to a “partner” means a person who is a partner, or equivalent, in such a law firm. Similarly, reference to an “office” means an office of any such law firm. This may qualify as “Attorney Advertising” requiring notice in some jurisdictions. Prior results do not guarantee a similar outcome.