Machine learning and artificial intelligence technologies (“data analytics”) are quickly transforming research and practice in law, raising questions of whether the law can survive as a vibrant profession for natural persons to enter. In this article, I argue that data analytics approaches are overwhelmingly likely to continue to penetrate law, even in domains that have heretofore been dominated by human decision makers. As a vehicle for demonstrating this claim, I describe an extended example of using machine learning to identify and categorize fiduciary duty waiver provisions in publicly disclosed corporate documents. Notwithstanding the power of machine learning techniques, however, I remain doubtful that data analytics will categorically displace lawyers from the practice of law, for two reasons. First, many of the most powerful approaches in data analytics as applied to law are likely to continue to require human practitioner inputs to train, calibrate and supervise machine classifiers. And second, the underlying evolutionary process that characterizes legal doctrine and precedent is irreducibly dynamic and complex – traits that are poorly adapted to pure algorithmic decision-making. Consequently, aspiring legal researchers and practitioners should not fear entering the field, but in doing so they would be well advised to invest in skill sets that are complementary to data analytics techniques.
Eric L. Talley,
Is the Future of Law a Driverless Car? Assessing How the Data Analytics Revolution Will Transform Legal Practice,
Journal of Institutional & Theoretical Economics, 174:1, pp 183-205, 2018
Available at: https://scholarship.law.columbia.edu/faculty_scholarship/2061