D.C. Bar Summit Explores How AI Will Revolutionize the Legal Profession
September 20, 2023
On September 18 the D.C. Bar gathered a wide array of experts in law and technology for its Artificial Intelligence & Chatbot Summit, a daylong discussion on the expected impacts — both positive and negative — of emerging technologies on the legal profession.
Keynote speaker William Eskridge Jr., Yale Law School’s Alexander M. Bickel Professor of Public Law, assessed attorneys’ current use of AI and chatbots and urged attendees to continue experimenting and educating themselves on the technology. Describing himself as a naysaying curmudgeon in the area of technology, Eskridge has, regardless of his reticence, been involved in AI experimentation, including overseeing its deployment in his law school classes.
In his experience, Eskridge said ChatGPT is capable of finding resources, locating information, making predictions, and providing analysis. “It can do a lot of legal reasoning,” he said. This is not breakthrough technology, Eskridge pointed out, citing TurboTax software and its ability to fit simple fact patterns into legal frameworks.
On the technology’s limitations, Eskridge said ChatGPT has a hard time handling analogy. “Even when [ChatGPT] finds relevant cases, it’s not always good at putting them together and reasoning by analogy,” he explained, describing his inquiry to ChatGPT regarding whether Title 7 protects nonbinary employees. ChatGPT was able to find a relevant case, Bostock v. Clayton County, but that case did not involve nonbinary employees, and ChatGPT struggled to apply the reasoning to an analogous situation that did.
Eskridge also noted that ChatGPT could answer questions about the application of a law, but it had difficulty discussing the law’s purpose or finding an interpretation of the legal text that addressed its purpose. “It’s a little bit guess-y,” he said. “Is it as good a job as a lawyer could do? Probably not, but it can do a lot of legal analysis.”
Given these strengths and weaknesses, AI tools can be very useful for conducting legal research, Eskridge said. “It doesn’t provide you the final answer, but it can give you a lot of useful information.”
In practice, AI could be helpful in making predictive judgments and other assessments that can guide litigation, according to Eskridge. Technology-assisted review (TAR) is already widely used in discovery to determine whether each document is relevant or discoverable, he said.
The largest changes predicted by Eskridge involve the structure of law firms, going from the traditional model with partners at the tip of a pyramid and associates and staff falling under their authority to one where partners are in the middle surrounded by technologists, outside consultants, and other attorneys.
This revolution in law firm structure would create a problem in jurisdictions where nonattorneys are barred from holding equity in a law firm, Eskridge said, but would give an advantage to the District of Columbia, the first jurisdiction in the United States to permit lawyers to form partnerships with nonlawyers in limited circumstances.
Partnering with a law firm that uses AI, Eskridge teaches a class at Yale conducted entirely on AI platforms. He said that the course demonstrates AI’s practical use in the field by attorneys and law clerks. Eskridge also sees law school clinics as another opportunity for the technology’s use in legal education, urging attendees to drive its adoption to address the underrepresentation of many litigants.
“Most people in lawsuits today have no attorney,” he said. “Maybe law schools can create mechanisms where more people can at least be informed about their rights in landlord–tenant court or be given documents that they can fill out themselves. I implore you all to work with [schools] to come up with innovative ways for lawyerless folks to get access to high-quality legal advice through these mechanisms.”
A key step in leveraging AI to expand access to justice would be loosening the rules relating to unauthorized practice of law (UPL) that constrain experimentation with AI systems.
Hope Todd, the D.C. Bar’s associate director of legal ethics, pointed out that amendments to UPL rules are under the purview of the District of Columbia Courts’ Committee on Unauthorized Practice of Law.
Meanwhile, in a later session titled “Where Do We Go From Here?” D.C. Bar Chief Programs Officer Darrin Sobin spotlighted the classes and working groups established by the Bar to educate both staff and members about AI tools. Sobin also said the Bar welcomes partnerships with area schools to innovate the provision of legal services through the use of AI technology.