P2P

79 I L T A N E T . O R G well as the business of law, in that to effectively manage legal operations, you need to budget appropriately for AI tools, and this includes the people that support them. You don't want that extra time and expense to come as a shock during a live matter or investigation. 4. We will need to create new roles. The increased incorporation of AI tools into discovery workflows might mean that we need to create entirely new jobs that weren't previously necessary. While it's true that advances in AI are going a long way toward expediting first-pass review, that's not the same thing as eliminating first-pass reviewers altogether. There's still plenty of work for humans to do. For instance, when we are dealing with portable models that need additional training or tuning and new data needs to be labeled and sorted before it can be reviewed, do we actually need junior attorneys to be performing those tasks? For example, with recent increased attention on protecting personal and private information, demand for AI-powered solutions that identify and extract private information is on the rise. There's an argument to be made that nonlawyers can easily annotate driver's license or credit card numbers or even that people with health care backgrounds are better suited to do an initial pass-through of sensitive health information. Acknowledging and creating the new roles that support the effective use of AI could free up attorneys who are currently handling some data-centric tasks to do different, higher-value legal work. In line with the current drive to be our best selves, we can allow our attorneys to be their best selves when technology lets them focus on the things that matter most to them and give them the greatest chance for fulfilling work. High-dollar attorneys should be working on high-dollar legal work, not data labeling or rote document review that requires no legal judgment. Moving Forward with a Realistic View of AI AI is not magic and it's not about to completely replace humans anytime soon, but it can have significant, beneficial effects on eDiscovery and isn't something to fear. If you've been thinking that your AI solutions have failed because they still involve people, you've been thinking about it all wrong. The key to successfully using AI – in eDiscovery or in any legal task – is understanding that humans will always be a part of the process. When considering AI solutions, be sure to ask questions and understand how and where humans will be involved. With these data points, you will better understand what AI really costs and how to best use it. Then, with the right people in the right roles, with the right skills to get the job done, AI can handle the rest. ILTA By Dr. Gina Taranto, ProSearch Director, Applied Sciences/Accelerated Learning Solutions, leads research and innovation of accelerated learning solutions by directing multidisciplinary teams of technologists, subject matter experts and data scientists to train the technologies that replicate human decisions. She has been developing teams and solutions in eDiscovery for more than 15 years, with experience in the design and implementation of search and automated document review solutions for clients in the financial services, technology and pharmaceutical industries. Taranto received her B.A. from Kresge College at the University of California, Santa Cruz, with honors, and her M.A. and Ph.D. from the University of California, San Diego.

• Cover