P2P

53 I L T A N E T . O R G minimization allows legal professionals to perform more thorough and meaningful e-discovery work. • Data Governance: Eventually, Gen AI is expected to work its way even further left of the EDRM into data governance, making activities around retention, security, reporting, and compliance more efficient. Using it from the start will also greatly impact downstream e-discovery activity in data output quality. The Ethical Considerations Though the benefits are clear, Gen AI in e-discovery comes with several ethical concerns—mainly because U.S. law lags in developing regulatory measures for the technology. Because of these gaps and the technology's nascency, using Gen AI opens the door to risk regarding ethics and defensibility. First, obvious privacy and security concerns are associated with collecting massive amounts of data. Instead of regulation, legal teams need to revisit their data governance programs and establish guardrails to ensure they operate ethically and lawfully when handling sensitive information. Gen AI is also prone to "hallucinating," meaning that it can output plausible, convincing, and sourced answers that are entirely inaccurate, which leaves companies and law firms liable for the error. Until Gen AI systems can offer 100% accuracy, comprehensive validation processes should be implemented into e-discovery workflows to minimize that risk. Finally, and most importantly, Gen AI is infamous for bias. Its outputs are only as good as its inputs, and all datasets include bias. Lawyers need to research how bias plays into and off their models and instigate controls to mitigate that bias. Regular bias audits, data augmentation to include diverse viewpoints, bias-focused validation techniques, and neural network bias training are all measures that lawyers can take to minimize unintentional harm. This work not only prioritizes ethical practices but also promotes defensibility. Preparing for an E-Discovery Future with Gen AI Companies and law firms should focus on a few critical foundational steps to prepare for a future in which Gen AI is a normalized part of the legal workflow, especially the e-discovery process. They should build out and stress test responsible, ethical AI processes with guardrails that uphold data privacy best practices, ensure accuracy, and mitigate bias. Second, they should evaluate Gen AI technologies to find one supporting these responsible practices and workflows. Lastly, despite being a powerful and largely autonomous technology, Gen AI requires the right people in the driver's seat, especially when setting up ethical workflows and tool selection. Legal departments and law firms must assemble the right teams (with principled, knowledgeable members) to guide this massive experiment and reduce risk along the way. By doing so, legal teams, especially in-house teams, can quickly expect unprecedented self-sufficiency and productivity. ILTA Brenda Dodd is the Director of Project Management for Exterro. Prior to Exterro, she spent five years at Reveal ZyLAB, as the Director of Customer Success and GTM Product Director. Brenda is an innovative Certified eDiscovery Specialist with over 25 years of legal experience in driving customer adoption and growth, leading product development projects, developing marketing strategies, and consulting clients regarding e-discovery and data management activities. Brenda is passionate about assisting law firms, corporations and government entities define best practices and procedures for cost-effective and efficient management of electronically stored information for legal review.

• Cover