The Other 'LLM': Large Language Models and the Future of Legal Education

Abstract

On 21 December 2022, the Dean of Suffolk University Law School, Andrew Perlman, posted an academic paper on the Social Science Research Network. This otherwise unremarkable occurrence was made remarkable by the identity of Perlman’s co-author: ChatGPT, a large language model (“LLM”) chatbot developed by OpenAI, an artificial intelligence (“AI”) research laboratory. ChatGPT wrote most of the paper; Perlman contributed the abstract and section headings and inputted various prompts. The results are good, not great. But the paper's key point is clear: disruption of the legal industry by LLMs and other AI applications is near-term, if not now. Legal education will not be immune to this disruption. In fact, in the first part of this paper I will argue that legal education will be one of the primary areas affected. Both law students and law teachers will feel the inevitable attraction of LLMs. Got an essay on negligence due tomorrow? No problem – get ChatGPT to draft a general outline, add in some case references, and voilà – an essay. Similarly, during a busy marking season, teachers may be tempted to ask an LLM to review student essays so that they can allocate marks based on those reviews. The responses to the LLM challenge will vary across law schools. Some will undoubtedly stick their collective heads in the sand by prohibiting ChatGPT. They may even be tempted to revert to handwritten examinations, mooting and in-class essays as the only assessment forms. In the second part of this paper, I will argue that such responses are misguided. After all, the challenge presented by LLMs is a challenge that other faculties have faced before – most notably, the mathematics department with the invention of the calculator. As such, I will propose that the usage of LLMs and other AI tools be legitimized in law schools. Such a step would challenge teachers to reimagine legal education as such a joyful, meaningful, and engaging enterprise that law students will want to do it all themselves – rather than let computers have all the fun.