How to Overcome the Challenges of Cross-Border eDiscovery

09
Sep
2013

Jonathan Shaman

When eDiscovery involves data from more than one country, it adds layers of complexity to an already complicated process. Hiring experts on the ground in each country, collecting documents from countries with privacy laws that differ from those of the United States, and dealing with documents written in unfamiliar languages are some of the many challenges counsel must overcome. 

Given the variations in law and cultural preferences, there is no one-size-fits-all model to handling cross-border eDiscovery. However, the following best practices can help bridge the legal, cultural, and linguistic divides.

1. Retain an Experienced Outside Support Team With Local Roots.

Communication plays a primary role in building relationships abroad, beginning with contract negotiations and spanning the entire eDiscovery process. To facilitate communication, language fluency is an obvious requirement; however, it is often better to employ native speakers to heighten trust. Native speakers can also alleviate the natural skepticism about outsiders that is present in many cultures. 

Therefore, the first step in managing a cross-border eDiscovery project should be assembling a seasoned team of local experts. 

2. Don't Get Lost in Translation When Identifying Documents.

Determining whether foreign custodians have potentially responsive data can be challenging, particularly when the custodians live in a time zone halfway around the world and do not speak the same language.  Similarly, dealing with clients' IT personnel in different regions often requires a deft touch as eDiscovery inquiries may be viewed as an intrusion.  In that case, the risk of misunderstanding requests for information and missing important documents is high. 

Translators fluent in the language can transcend the language barrier. Consider hiring an eDiscovery specialist with native speakers on staff who can interview custodians and who have the technical acumen and sensitivity to facilitate discussions with IT personnel, adding defensibility to the identification process.

3. Avoid Privacy Pitfalls During Collection.

When collecting data abroad, counsel must understand all applicable privacy laws.  The need to transport information often conflicts with the legal restrictions on the movement of personal data at both national and supra-national levels, which are imposed by regulations such as the European Data Privacy Directive and the Japanese Personal Information Act of 2003. Some legal instruments concerned with the protection of corporate information, including the French Blocking Statute of 1968, even criminalize data movement in some circumstances. State data restrictions such as the Law of the People's Republic of China on Guarding State Secrets can require the state to review certain classes of documents before their movement outside the original jurisdiction. Key implications of these hurdles include requirements to notify the individuals concerned, to obtain consent and to minimize the amount of personal data moved prior to review by using filtering technology, either in the country of origin or in another approved jurisdiction. 

Because privacy rules vary, outside counsel and eDiscovery experts must understand local rules and regulations governing the collection, processing, and transport of data. Counsel may determine that it is safer to send collection or review teams to the deal with the data in-country rather than exporting it to the United States. The client's eDiscovery consultants should offer a number of options - both technical and human -- to assist counsel in determining the best solution. 

4. Choose a Platform That Can Homogenize Data for Filtering and Review.

After data is collected, it must be processed and uploaded into a document review platform. The application must provide an efficient way to review documents that contain a number of languages, to categorize and prioritize them for review, and to apply attorney work product indicating responsiveness and privilege, redactions for privileged and sensitive information, and comments.  Of equal importance, the software solution must ensure the security and confidentiality of the documents, and comply with the restrictions required under applicable laws.

As an initial matter, the eDiscovery software must be able to bring together various formats and file types, including documents derived from email, native files, backup tapes, and legacy data, and even audio or video recordings, in a single global review platform. This can be especially taxing when dealing with documents in different languages encoded in non-Unicode formats. If the review platform is not flexible enough to handle these files, then reviewers must undertake the time-consuming and costly task of reviewing documents in their native formats or in hard copy. 

Additionally, to keep costs in check, the platform should incorporate tools capable of reducing the data volume to a manageable level, such as deduplication, and technology assisted review tools like concept clustering, threading and predictive coding. To accomplish this reduction defensibly, the platform must be able to recognize, filter, group similar documents, and search for numerous letters, numbers, and other characters. This process can become tricky for languages such as Chinese and Japanese that do not use spaces to mark the beginning and ending of words.  It is important to understand how the review software handles these languages to determine the best course of action.

5. Evaluate the Merits of Human vs. Machine Translation.

In the best of all possible worlds, non-English languages would be translated by humans who understand all the nuances of both the source language, and English.  Unfortunately, human translation is time-consuming and expensive. Machine translation is less costly and more efficient than human translation, but it sacrifices accuracy. Combining human and machine translation often yields the best results. An initial pass with machine translation can quickly determine relevance and assist in culling out irrelevant documents. Then, human translators can focus on a narrower, prioritized set of documents. Furthermore, in examples such as high profile government investigations, the company's investment in human translation can express a sincere concern to resolve the conflict without extending the discovery timeline for months while human translators waste time reviewing nonresponsive documents.

To avoid the expense of hiring domestic lawyers who are fluent in non-English languages, engage a service provider with access to local, native-speaking reviewers who have experience in document review. This approach alleviates the headaches of managing the complex HR components of recruiting, onboarding and paying foreign contractors. In addition, the service provider can take care of logistical arrangements to maximize the project's efficiency, such as procuring office space, computer equipment, and software.  Service providers can also be helpful in navigating local cultural nuances in training and management of these review resources.

6. Utilize Technology Effectively to Control Costs.

Technology to assist the review holds great promise at reducing costs and finding responsive documents.  But, when working with data sets collected across country borders, the technology introduces special nuances and considerations that counsel should consider when constructing defensible and efficient workflows. The key to using eDiscovery tools effectively is understanding their appropriate uses, and their strengths and weaknesses - and the challenges that might need to be overcome when working with complex data sets. 

For example, traditional search technology is tried and proven for culling and review prioritization, but since search queries are language-specific, counsel may have to hire a translation service in order to construct the query set to enable search against a multi-lingual document corpus.  And, search terms that are indicative of individual lawyer or actor involvement, such as the names of corporations, law firms and counsel, as well as variations of individual email address might need to be expanded when the data set contains multiple languages, countries and corporate subsidiaries.

Predictive coding as a classification technology is also a very powerful tool that can reduce the reviewed document volume by eliminating certain non-responsive documents.  But when working with multi-lingual data sets, counsel must be sure to select a predictive coding engine that is proven in those circumstances - handling right to left written languages, or data sets that contain character based or Cyrillic characters.  Just being built on a Unicode-compliant architecture is not enough.  Counsel should also consider whether the predictive coding classifier engine is proficient at modeling individual documents that contain multiple languages, or whether the data sets needs to be separated by language for the modeling engine to work effectively. 

Conclusion

When handling global discovery projects, choosing the right combination of people, process, and technology can be the difference between success and failure. To maximize the defensibility of projects, organizations must leverage the expertise, language skills, and cultural knowledge of local eDiscovery experts; embrace adaptive workflows that respect privacy laws; and match technology to client needs and expectations.