In litigation and government investigations, lawyers have a duty to disclose all documents related to the dispute, no matter where that data lives: whether it’s on a laptop, on a smartphone, across multiple servers, or in a cloud-based database. Therefore, this information explosion has transformed what was once a challenge akin to walking up a steep hill into hiking Mount Everest. For example, in a recent lawsuit it filed, the U.S. government buried the defendants’ lawyers in over 250 terabytes of data (enough paper to store in the Library of Congress about 20 times over).
A CIO or general counsel tasked with handling the document requests during a lawsuit understands how unwieldy it is to provide years of company data, consisting of large amounts of information, during the discovery process without the right resources. That’s where automation has stepped in. The latest breakthrough has come to be known as “predictive tagging,” which essentially teaches the e-discovery software what types of documents are relevant to each case. Complex algorithms help the program learn to “review documents,” marking those documents for responsiveness or privilege.
Will this technology effectively replace all human document reviewers? The short answer is simply no. Despite all the technological advances in e-discovery and document review, without human guidance, machines cannot handle the complex decision making that ultimately determines whether a document is responsive. Lawyers must first manually go through a sample set of documents and teach the program what is relevant and what is not. Once the software has learned what to look for, legal teams can benefit from the software’s quick prioritization of data. But ultimately, it is still the lawyer’s job to confirm the results.
Predictive tagging is particularly indicative of how automation and human comprehension must collaborate to efficiently complete the task of reviewing data. Essentially, they need each other. Because humans have to teach software how to “understand” and therefore treat certain data, people are an indispensable part of the process. Yet, without the accurate culling that software performs, the amount of data would be too large for humans to manage and review alone. As an added benefit, some software and e-discovery providers can monitor the speed, progress, and accuracy of the human reviewers to determine their efficiency and competency.
That’s where automation reaches the boundary of practicality. Indeed, software is brutally efficient; it doesn’t fatigue, and it never makes a mistake. Then again, it cannot stray off a given course. Therefore, human intuition, awareness of nuances and ingenuity will always need to play a part in the process.
via Balancing Automation And Hands-On Review In E-Discovery – Forbes.
