Case Details
- Title: Robin Duane Littau v Astrata (Asia Pacific) Pte Ltd
- Citation: [2011] SGHC 61
- Court: High Court of the Republic of Singapore
- Decision Date: 18 March 2011
- Case Number: Suit No 156 of 2010 (Summons No 5244 of 2010)
- Tribunal/Court: High Court
- Coram: Yeong Zee Kin SAR
- Plaintiff/Applicant: Robin Duane Littau
- Defendant/Respondent: Astrata (Asia Pacific) Pte Ltd
- Counsel for Plaintiff: Gan Kam Yuin, Pua Li Siang and Joanna Lee (Bih Li & Lee)
- Counsel for Defendant: Gerald Kuppusamy and Shaun Lee (Wong & Leow LLC)
- Procedural Context: Application under Part IVA of the Supreme Court Practice Directions for discovery and inspection of electronically stored documents (“Part IVA”) to determine keywords for a reasonable search after execution of an Anton Piller order
- Legal Areas: Civil Procedure; discovery of documents; electronic discovery; Anton Piller orders; disclosure/production of documents; privileges
- Judgment Length: 15 pages, 8,247 words
- Prior Related Decision Adopted: Littau Robin Duane v Astrata (Asia Pacific) Pte Ltd [2010] SGHC 361 (adopted for personalities, issues and procedural history)
Summary
This High Court decision concerns the practical mechanics of electronic discovery following the execution of an Anton Piller (search) order. The plaintiff, Robin Duane Littau, had been subject to interim injunction and search orders obtained by Astrata (Asia Pacific) Pte Ltd in the context of a dispute over alleged misuse of confidential information and breach of a non-compete agreement. After the search orders were executed and forensic images of the plaintiff’s computers and email accounts were seized, the parties could not agree on the keyword list and the procedure for conducting a “reasonable search” for discoverable documents. The court was therefore asked to determine the keywords to be used under Part IVA of the Supreme Court Practice Directions.
The court held that the relevance test for determining the scope of documents to be captured by keyword searches in the execution of search orders is not meaningfully different from the relevance test applicable to discovery in the pending suit. Given the draconian nature of Anton Piller orders, the executing party must meticulously comply with the limits of the order; “95% accuracy” is not sufficient. Accordingly, the keyword search must be anchored to the issues in dispute and must be designed to identify documents that are relevant to those issues, while also respecting exclusions such as litigation privilege.
What Were the Facts of This Case?
The plaintiff, Robin Duane Littau, was formerly a group regional director of Astrata (Asia Pacific) Pte Ltd. Between March 2005 and December 2009, he worked in senior executive and managerial roles in the Asia-Pacific region. During his employment, he signed a Confidentiality, Restrictions & Intellectual Property Agreement (the “non-compete agreement”). On 11 September 2009, he gave three months’ notice of resignation, with his last day of employment expected to be 10 December 2009. However, on 9 December 2009, Astrata purported to dismiss him summarily with immediate effect.
The plaintiff commenced proceedings in relation to unpaid salary for the period from September to 10 December 2009, as well as expenses and accrued leave. Astrata filed a defence and counterclaim alleging that the plaintiff breached his duties by giving information to certain entities (the “US parties”), which allegedly caused Astrata to incur extra costs in defending a Chapter 11 proceeding in the United States. Shortly before Astrata sought interim injunctive relief and search orders, it sought leave to amend its defence and counterclaim to include a new claim for breach of the non-compete agreement. The amendment was allowed, subject to limitations on the use of evidence obtained in the Singapore action in other proceedings involving the Tridex companies and Astrata’s group.
On 21 June 2010, Astrata obtained an interim injunction and search orders against the plaintiff. The injunction restrained him from giving confidential information to third parties and from breaching the non-compete agreement. The search orders were directed at evidence that the plaintiff had given information to the US parties and at evidence of breach of the non-compete agreement. The search orders were extended because the physical imaging of the plaintiff’s Gmail account could not be completed in time. A moratorium was ordered against perusal of materials obtained, with the prohibition continuing until further notice if the plaintiff applied to set aside or vary the orders.
After the moratorium was lifted, the parties were directed to review the disputed documents and attempt to agree whether they fell within the ambit of the search orders. The seized items included forensic images of the contents of the plaintiff’s Apple iMac desktop hard disk, his PowerBook notebook hard disk, and the contents of his Gmail, LinkedIn, and AOL email accounts. The forensic imaging produced a sizeable volume of data, making manual review impracticable and expensive. The parties agreed that keyword searches should be used to identify relevant documents, and they also agreed on the use of keyword searches to identify documents subject to litigation privilege for exclusion. However, they could not agree on the list of keywords or on the procedure to be adopted after keyword searches yielded responsive “search results”.
What Were the Key Legal Issues?
The central issue was how to determine the keywords for a “reasonable search” for discoverable documents under Part IVA, following the execution of an Anton Piller order. Specifically, the court had to decide whether the relevance threshold for keyword searches conducted under the search order regime should differ from the relevance threshold for discovery in the pending suit. This question mattered because the parties disagreed on whether the executing party’s powers under the search orders should be broader and less constrained than the powers to conduct discovery.
A second issue concerned the proper approach to electronic review in a high-stakes context. The court needed to ensure that the keyword search process would not become an overbroad mechanism that effectively expands the scope of the search order beyond what is permitted. This required attention to the draconian nature of Anton Piller orders and the need for meticulous compliance with their limits, particularly where the search involves large volumes of personal and confidential electronic data.
Thirdly, the court had to consider how the keyword search should operate in a structured review framework. The parties had already agreed on a three-stage approach: (1) agree on keywords and conduct a reasonable search on the forensic images; (2) review the search results to identify confidential information and information subject to privilege; and (3) incorporate non-privileged, necessary documents into general or further discovery. The present application focused on the first stage—determining the keyword list and the relevance-based constraints that should govern it.
How Did the Court Analyse the Issues?
The court began by situating the application within the broader procedural history and the established review framework. It adopted the earlier account of the parties’ dispute and the circumstances leading to the search orders. It then addressed the impasse on keywords and the legal test for relevance. The court emphasised that ocular examination of the electronic documents was not feasible due to the volume of data and the cost of review. Keyword searching was therefore the practical and preferred method to identify potentially discoverable documents within the forensic images.
On the question whether the relevance test for keyword searches under search orders differs from the relevance test for discovery, the court rejected the defendant’s attempt to draw a distinction. The defendant argued that its powers under the search orders were broader and should not be fettered by a list of keywords tied strictly to issues in dispute. The plaintiff, by contrast, argued that keywords must be relevant to the issues in dispute and that responsive documents would then be reviewed for final disclosure. The court held that it did not see any meaningful distinction between the test for relevance for compliance with the search orders and the test for relevance for discovery in the pending suit.
In reaching this conclusion, the court underscored the exceptional and “draconian” character of Anton Piller orders. Such orders are granted only in exceptional cases because of their drastic effect and far-reaching consequences when executed. The primary purpose of a search order is to preserve evidence relevant to the case from destruction or concealment where there is solid evidence of a real risk of such destruction. This purpose is not served by allowing the executing party to cast a wider net than the order permits. The court therefore treated relevance as a gatekeeping requirement: documents within the scope of the search order must first be relevant to the pending suit in which the order was made.
The court also linked this relevance requirement to the obligation of the party executing the order to ensure full compliance with its scope. It cited the principle that “95% accuracy” is not good enough in this context. Because Anton Piller orders can become instruments of oppression if their limits are not meticulously obeyed and enforced, the executing party must ensure that the search and seizure process does not exceed the order’s boundaries. This strict approach applies not only to the initial seizure but also to subsequent steps such as keyword searching and review, because those steps determine what documents are effectively brought into the discovery process.
Against that background, the court treated the keyword list as a mechanism that must operationalise the relevance requirement. Keywords cannot be selected in a way that effectively authorises a general rummaging through electronic data. Instead, the keywords must be designed to identify documents that are relevant to the issues in dispute—namely, the alleged provision of confidential information to the US parties and the alleged breach of the non-compete agreement. The court’s analysis therefore required a careful alignment between the keyword list and the substantive allegations that the search orders were intended to capture.
Finally, the court addressed the procedural structure for electronic review. It had previously directed a three-stage framework: keyword search, followed by counsel review for confidentiality and privilege, and then incorporation of non-privileged, necessary documents into discovery. The court’s reasoning indicates that the keyword stage should not be used to bypass the later privilege and confidentiality safeguards. Rather, the keyword stage should be limited to producing a manageable set of responsive documents that can then be properly reviewed for privilege and relevance to disclosure.
What Was the Outcome?
The court granted the application and determined the keywords to be used for the reasonable search following the Anton Piller order. In doing so, it applied the principle that the relevance test for the keyword search must match the relevance test for discovery in the pending suit, and it rejected any approach that would broaden the scope of the search order beyond its intended limits.
Practically, the decision provided a legally constrained method for electronic discovery in a post-search-order setting. It ensured that the parties’ review could proceed without an open-ended keyword list, and it reinforced that the executing party must meticulously comply with the scope of the search order to avoid oppression and overreach.
Why Does This Case Matter?
This case is significant for practitioners because it addresses a recurring and difficult problem in modern litigation: how to conduct electronic discovery after a search order has been executed, when the volume of data is enormous and manual review is impractical. The court’s insistence that relevance for keyword searching must align with relevance for discovery provides a clear constraint on the use of keyword tools. It prevents keyword searching from becoming a de facto expansion of the search order’s scope.
From a precedent perspective, the decision reinforces the doctrinal foundation for Anton Piller orders in Singapore: they are exceptional, their purpose is evidence preservation, and their limits must be meticulously obeyed. By emphasising that “95% accuracy” is not sufficient and by treating relevance as a strict gatekeeping requirement, the court strengthens the procedural safeguards against oppression. This is particularly important where the search involves personal email accounts and other sensitive electronic repositories.
For lawyers, the case also illustrates the value of structured electronic review frameworks under Part IVA. The three-stage approach—keyword search, privilege/confidentiality review, and then discovery—helps ensure that privilege protections are not undermined by the initial search process. Practitioners seeking search orders or defending against them should therefore pay close attention not only to the wording of the search order itself, but also to the downstream mechanics of keyword selection and the review process.
Legislation Referenced
- Supreme Court Practice Directions (Singapore) – Part IVA: Discovery and inspection of electronically stored documents
Cases Cited
- [1989] SGHC 98
- [1991] SGCA 28
- [1998] SGCA 60
- [1999] SGCA 28
- [2002] SGHC 247
- [2006] SGCA 1
- [2007] SGHC 133
- [2010] SGHC 361
- [2011] SGHC 61
Source Documents
This article analyses [2011] SGHC 61 for legal research and educational purposes. It does not constitute legal advice. Readers should consult the full judgment for the Court's complete reasoning.