ARBITRATION WORLD The Relative Rewards and Risks of Predictive Coding Julie Anne Halter, Rob Noreus, and Mike Goodfried, K&L Gates e-Discovery Analysis & Technology Group As personal computers, laptops, tablets, smartphones, social media, cloud computing, and other devices proliferate in the workplace, so has the number of different software applica- tions that allow us to communicate from anywhere, at any time, and in virtually any format. As a result, the amount of information businesses create has exploded. Over the next 15 years, the digital universe is expected to more than double every two years, going from 4.4 trillion gigabytes to 44 trillion gigabytes. While generating and sharing information is critical for any business, one unexpected yet significant consequence of these enormous data volumes is the increasingly prohibitive cost of collecting, filtering, reviewing, and producing massive volumes of electronically stored information (ESI) in dispute resolution proceedings. Electronic discovery commentators estimate that at least 50 percent of ESI maintained by a given organization is duplicative, outdated, or unnecessary for business purposes. Accordingly, when faced with legal proceedings, clients often find that most of the ESI they collect is irrel- evant. The challenge for lawyers is how to best leverage rapidly evolv- ing technologies to reduce the time and expense associated with docu- ment review, while not compromising their duties as zealous advocates for their clients. Some document review solutions have been available for many years, but are enhanced with better technology. For example, the use of key words to search for and capture potentially relevant electronic docu- ments, and leave aside the vast majority of likely irrelevant documents, has been prevalent for more than a decade. Newer technologies allow 84 K&L Gates: ARBITRATION WORLD