The evolution of technological solutions to collect and analyze data continues to enhance the ability for businesses to prevent, investigate, and defend against allegations of business crimes
In a two-part webinar series, BLG’s Investigations & White Collar Defence Group explores data-driven risk management in terms of compliance, investigation and defence.
Summary of session 2
Presented originally on November 10, 2021.
Presenter: Graeme A. Hamilton, Partner at BLG Toronto and National Co-Chair of BLG's Investigations and White Collar Defence Group
Panelists:
- Laurie Goldbach, Partner at BLG Calgary and National Leader of BLG's Tax Litigation and Disputes Group
- Kelly Friedman, Senior Counsel & National Leader, Beyond eDiscovery
With the advent of smarter, more efficient tools to collect and organize information, data analytics has changed the conduct of investigations. The panelists discuss:
- How document review practices have evolved to incorporate data analytics;
- How government regulators use similar tools; and
- The impact of data analytics on defence lawyers and the expectations now placed on targets of an investigation, for instance in response to a search warrant or a production order, and issues of privilege that may arise.
The webinar concludes with a short demonstration on the application of data analytics tools.
The evolution of document review practices
A few decades ago, conducting a document review involved sifting through dozens of banker boxes and trying to create a chronology – a monumental task. In the early 2000s, as emails and other electronic documents became relevant, practices adapted. Initially, this adaptation included forensic examiners imaging hard drives, phones, servers and other electronic equipment. Although there was a gain in variety of the types of data, there was very little flexibility: the documents collected were all in picture format, preventing interaction with the underlying data available in the source document.
Over the last two decades, electronic communications have increased ten-fold, with the pandemic exacerbating our reliance on electronic communications. Between emails, text messages and chats from various messaging platforms such as MS Teams, Zoom and others, not only is electronic data more plentiful than ever, but it also allows for remote collection of this data, without necessitating access to hardware.
Given the enormous volume of data at hand, the key to processing the data efficiently is to have a robust discovery platform that incorporates artificial intelligence (AI). Current-day electronic discovery tools convert documents from disparate sources into a format that allows them to be analyzed together and easily cross-referenced.
Nowadays, AI-powered document reviews extend well beyond searching for keywords. Using conceptual analytics, they now feature:
- Keyword expansion, which involves looking into a dataset for conceptually similar terms to the ones fed to an eDiscovery tool, not just synonyms.
- Clustering, meaning that a review tool can also group together into packets conceptually similar documents within a dataset, based on initial findings.
- Name normalization, which combines personal and professional online aliases into a single, convenient entity that is easier to trace.
- Communication analysis, which visually depicts who is speaking to whom, and that can be modulated to reflect specific topics or moments in time.
The use of eDiscovery tools by government regulators
To date, the use of data analytics by government regulators varies by agency. For example, the Canada Revenue Agency (CRA) uses machine learning to develop predictive analytics, which identify potential areas of tax noncompliance.
At the audit level, these tools uncover unseen data patterns. At the investigation level, the CRA is also developing new data models to identify high-risk taxpayers, analyzing their social networks and automating the identification of links between individuals and businesses. The CRA has also announced that it will use these tools to combat offshore tax evasion.
The use of data analytics has allowed the CRA to bolster its evidence to readily obtain a search warrant. Despite the availability of remote search warrants, the CRA still conducts on-site search and seizures. Defense counsel must consider that a physical seizure requires immediate attention to privileged information in the seized materials.
The impacts on defence counsel and investigation targets
Waiver of privilege is one of the primary concerns for defence counsel in the context of an investigation. The increase in information spurred by electronic communications has created serious challenges for Canadian courts’ and regulators’ respective processes to assess privileged claims. Due to the large number of communications, the typical methodology of assessing a claim of privilege for each communication is no longer practical.
In order to demonstrate claims of privilege over large data sets, defence counsel may rely on analytical tools. Counsel must demonstrate to the court that there is a reasonable process in place to identify privileged communications. For example, parceling data into sets where the same privilege claim applies and the same elements used to designate a communication as privileged are present. In reviewing a few communications, a court may rule that any communication with those same elements present are also privileged and properly excluded from a production.
A prudent approach to conducting an investigation, or responding to a regulator’s search warrant or production order, is to contact legal counsel to assess the best method for collecting responsive communications and data. Counsel must assess the scope of the production and the sources of responsive information, assess appropriate privilege claims, and execute the privilege claims (saving privileged records in distinct folders, labeling them as privileged).
Counsel should remain present throughout the execution of an on-site search warrant, and should provide parameters for the execution of a remote search warrant.
