In some trade secret, unfair competition and noncompete cases, where the amounts at stake are significant, i.e., at least well into the seven or eight figure range, it comes as no surprise to anyone that the volume of potentially relevant electronic documents that will need to be searched and produced will be quite voluminous, with the numbers of documents occasionally surpassing the actual amounts of money at stake.
At the risk of stating the obvious, this presents an enormous logistical challenge, because the sheer costs of eDiscovery alone threaten the viability of otherwise legitimate claims and appropriate affrimative defenses to these claims. Simply put, the cost in legal fees of searching hundreds of thousands, or millions of documents one by one is cost-prohibitive to all but the largest institutional clients, and even for them, the value of that time relative to the certain cost is rarely worthwhile.
Similarly, while specific word searches are extremely useful, they are far from perfect; a slight change of verbiage can easily miss an otherwise critical set of documents.
Enter AI.
To be sure, AI is, at first blush, a massive upgrade over a simple word search, as it can also identify similar words and phrases, and catch more pertinent documents without incurring massive man-hours to do so. That said, it is not - at least not yet - a magic bullet, a one-stop solution to all the foregoing problems, as it does not come without huge red flags of its own. Simply put, AI is not infallible, and it too can miss things, even if it is far better than a simple word search. Additionally, it too isn't cheap - albeit still a very small fraction of the legal fees that would otherwise be incurred for conducting the searches.
