In line with the explosion of generative pre-trained transformer (GPT) artificial intelligence in the corporate world, Stradley Ronon Stevens & Young recently designed and implemented programming, training and iterative feedback loops for lawyers and business professionals across the firm to enhance the knowledge, confidence and real-world applications of GPT-driven tools. This group, known as the GPT Frontrunners, is unique in its focused, high-touch approach to curious, ethical engagement with generative AI.
Recognizing that generative AI can be a ”force multiplier” for midsize firms that are willing to embrace it, Stradley Ronon’s GPT Frontrunners are engaged in understanding the evolving opportunities and limitations, and educating the firm’s lawyers and clients accordingly. A recent LexisNexis report on generative AI in the legal profession found that 67 percent of corporate counsel expect their law firms to use the technology, and lawyers and legal professionals have an ethical obligation to understand the benefits and risks of relevant technology. There is little doubt that generative AI will play a role in successful, modern law firms.
Stradley Ronon’s reputation as a powerhouse full-service law firm that punches above its weight is borne out in initiatives like the GPT Frontrunners. Consisting of 35 lawyers and business professionals across the firm’s practices and departments, the team volunteered to learn about and interact with generative AI programs, assessing the benefits and risks of using such tools in the legal environment, and exploring how the technology could be leveraged to increase efficiency and fulfill client expectations.
“Taking a nimble, people-centric approach to leveraging technology, we are committed to giving our lawyers and business professionals early exposure, training and tools to learn how to speak and work with generative AI,” Chief Operating Officer Marissa Parker said. “Our training around prompt engineering is a terrific example of practical and foundational know-how in this space, equipping our teams to find smart and secure opportunities to bring generative AI efficiencies directly to our clients.”
Over the course of five months, the group — led by Patrick Arkins, the firm’s manager of business technology solutions and innovation — experimented with the most prominent, accessible GPT platforms. Focusing primarily but not exclusively on ChatGPT, the group conducted research and trials, attended trainings and conferences, and developed internal training sessions. Educated on data privacy, confidentiality and the very real dangers of hallucinations, the team found ChatGPT was most useful for jump-starting creativity and problem-solving business solutions, generating outlines, answering non-legal technical questions, and quickly amassing knowledge on key terms and acronyms in new subject areas. ChatGPT assisted the team with creating project proposals and timelines and improving the efficiency of processes that move data from one system to another. On the other hand, summaries of publicly available legal documents were unreliable at a high rate, as were responses to probing questions about legal analysis.
Clients have already benefited from the group’s research: In one case, the team used ChatGPT to brainstorm techniques, incentives and penalties a client could employ to maintain control over the timing of a complex development project. In another, Arkins worked with a client to expedite the crafting of a job description for a newly created role. A few short discussions, combined with information provided by ChatGPT, produced a solid outline of responsibilities and qualifications in minutes. Key to these successes was Arkins’ deep knowledge of prompt engineering: the way in which humans talk with GPT tools.
Arkins and the group hosted CLE trainings aimed at educating the firm’s lawyers and business professionals on generative AI and best practices — sparking discussions in the process on how the firm can best utilize these tools in specific-use cases and more broadly. The group is offering lawyers assistance with AI tools for approved projects after they complete required training. Stradley Ronon is also experimenting with developing its own chatbots and exploring the potential benefits of both building its own secure GPT system and a GPT system that can access and utilize the firm's own documents and data as an information source.
“Now that we have lessons learned from our GPT Frontrunners program, an understanding of the risks involved, policies and guidelines to mitigate those risks, training resources and a bank of useful prompts, we are ready to provide these tools to any of our lawyers and business professionals who have a valid business need,” Arkins said. “We will continue experimenting and learning with freely available tools, developing necessary use cases to leverage those tools, and determining where to invest our time and hard-dollar resources in more advanced solutions.”
Information contained in this publication should not be construed as legal advice or opinion or as a substitute for the advice of counsel. The articles by these authors may have first appeared in other publications. The content provided is for educational and informational purposes for the use of clients and others who may be interested in the subject matter. We recommend that readers seek specific advice from counsel about particular matters of interest.
Copyright © 2023 Stradley Ronon Stevens & Young, LLP. All rights reserved.