hroughout the eDiscovery process there are critical moments where mishandled data or sloppy analysis could derail the entire effort and result in disastrous outcomes. The EDRM model is our best attempt to develop a coherent series of standards that ensure industry alignment, optimized results, and a common language. However, the EDRM model and its individual stages also helps to prevent problems from occurring along the way. These standards help us understand not only what we are doing and why, but also serve a vital risk management function. As the industry has grown, these standards have continued to improve.
In the legal world, risk is everywhere. The operational skill required to maintain a high level of standards is pivotal. This is a never-ending challenge that can make or break cases, impact relationships, and ultimately speak to a firm’s reputation. It is fair to say that the eDiscovery industry has little tolerance for error. With this in mind, it makes sense that each firm demand’s rigorous testing before they implement a new technology into their existing workflows. When considering the application of a new solution, the firm must be sure it is the right decision. Legal technology vendors are creating new features every day. Each year trade shows scale to larger venues to accommodate the new providers entering the market to solve the myriad of challenges within the eDiscovery industry. Yet for any of these tools to find oxygen, eDiscovery professionals must be confident that they will not increase their risk profile.
Today there are a number of new challenges which have increase the dimensions of risk within the eDiscovery industry. The explosion of data, in both type and size, has only served to enhance the importance of utilizing artificial intelligence (AI) platforms to cleanly and efficiently reduce large datasets to most relevant documents. But this isn’t the only challenge we face. The distractions caused by civil unrest have only increased in the past decade. Focusing on complex investigations and data analysis is difficult enough in a peaceful world, let alone one in which the general public feels divided. This year the distractions only grew when the transfer of entire workplaces to remote home working environments created an entirely new set of challenges for maintaining employee productivity and engagement. These range from managing the family environment and virtual schooling schedules, to mental health challenges through increased stress and anxiety.
These are very human issues, and despite our best efforts they don’t appear to be abating. For the eDiscovery industry this presents a unique threat to the quality of the finished product. For years there has been arguments for increasing the application of digital aides in order to supplement the basic functional challenges that prevent people from working as efficiently as robots. We need to eat, sleep, unwind, maintain healthy social lives, enjoy days off with friends and loved ones, and find meaning in our existence beyond that of a battery. However, with no clear end-date in sight for the travel restrictions and health protocols that have been implemented since March, it’s time to accelerate the application of AI tools and reduce the risk of human error.
Litigation teams already utilize AI technology for navigating large data sets in ways that prove impossible for human workers. Electrical circuits process information around a million times faster than biochemical ones. It is our general intelligence that ensures our superiority at the task of solving complex problems. The purpose of these tools is for removing the mundane analysis and excess hands within the document review process. The more people that are involved, the greater the risk of something going wrong. This is the eDiscovery equivalent of there being too many cooks in the kitchen. Instead of using fancy concepts and over-staffing eDiscovery workflows, there is a vital need to simplify and let AI do the heavy lifting.
We need to consider our humble shortcomings and employ AI to reduce the risk of human error throughout the review process. This can be achieved by implementing a Fordian process. The object should be to take a law firms DNA and turn it into a library of AI models. The models can then be catalogued by industry, practice group, and the relevant stages of the eDiscovery workflow. The key is the quality of the algorithms used to build the models, and access to suitable training data. Thankfully, law firms have no shortage of data. When a new matter arises, the firm can simply deploy the models at the start of the review process and reveal all relevant and privilege documents.
By turning legal data into AI models, organizations can ease the strain on the individuals they employ. This must start with the review stage and carry on to more tailored applications such as identifying sensitive data entities, practice specific insights, and proactively monitoring for opportunities to litigate. At the end of the day, we’re measuring the risk of people getting things wrong in traditional document review vs the risk of a well-managed AI workflow. The robots are here to help us come up for air.