The legal writing classes required of first-year law students tend to polarize students. Some love legal writing, deriving pleasure from the highly formulaic yet eminently adaptable paradigms of legal analysis. Others loathe it, feeling that it stifles creativity, imposing a repetitive and cumbersome structure on what might otherwise be a compelling and concise argument. As a law student, I fell into the former category, but when I became a legal writing professor, I found myself somewhere between the two.
When ChatGPT was released to the public in November 2022, it was similarly polarizing—at least among academic administrators and instructors. Most responded to its arrival with dismay. Renowned linguist Noam Chomsky opined that ChatGPT doesn’t have “anything to do with education except undermining it.” Many school systems and at least one top-tier law school immediately adopted policies to ban its use on written assignments and exams, at least until it was better understood. Some law professors who historically gave open-book final exams expressed grave concerns that students could use ChatGPT to produce decent, if not good, exam answers. Some were more ambivalent, and a minority hailed ChatGPT as a robust tool that could reduce the tedium of law practice. Nearly all feared that students would grow to rely on ChatGPT rather than learn the critical skills of legal analysis and writing. Many educators whose institutions did not globally prohibit ChatGPT did so in their own classrooms and carried on as usual. Others ignored it altogether. I—and a few others—embraced the nascent technology and quickly began discussing it in the classroom.
As someone who fell in love with diagramming sentences in seventh grade and fostered that romance for over a quarter-century to become a writing instructor, I viewed ChatGPT with excitement and trepidation. In ChatGPT, I saw potential. I saw a tool that could almost effortlessly merge the science of technology and art of writing. But most importantly, I saw an opportunity to teach my students not just how to write, but how to adapt to the rapidly changing landscape of legal practice.
The legal profession, which at its core is averse to change, is not immune to the digital revolution. From e-discovery tools to legal research platforms, technology has already transformed the way lawyers practice law, and much of that transformation has occurred within the past two decades. The challenge for legal educators is to prepare students for this new reality while still teaching them the fundamental skills they need to be effective lawyers.
ChatGPT, I believe, is part of this new reality. For that reason, I requested (and was approved) to teach a summer course at Rutgers Law School on using ChatGPT in law practice. The two-week intensive “skills” course covered the fundamentals of how ChatGPT works (and the essential vocabulary, like AI, ML, NLP, and LLMs), the strengths and weaknesses of the tool as applied to specific tasks within law practice, and the ethical considerations—with an emphasis on developing the ability to assess the suitability of any AI-driven tool for use in a legal practice setting. In short, the class was a success.
The classroom buzzed with enthusiasm and curiosity, but also with puzzlement that this was, to date, their only opportunity for guided instruction on a tool that was predicted to revolutionize the practice of law. By completing tasks of increasing difficulty under time constraints that would make those tasks nearly impossible for law students (or anyone lacking subject-matter expertise), they developed proficiency using ChatGPT for many types of tasks lawyers do on a daily basis, such as drafting and editing documents and conducting research. Through discussions that focused on processes rather than outcomes, they learned how to break down complex tasks into their constituent sub-tasks and identified which of those sub-tasks were well suited for ChatGPT. They learned to discern tasks based on their suitability for ChatGPT and began to view ChatGPT as akin to the other tools that routinely receive instruction in law school, like Westlaw, LexisNexis, and other subscription databases.
Is this outsourcing fundamental lawyering skills to a machine? For students properly instructed, I don’t believe so. If anything, it’s teaching students how to leverage technology to improve their workflows and use the tools at their disposal to be more productive and efficient. With AI-driven tools expanding and evolving in every sector, including the legal industry, these are valuable skills that will prepare them for the future of legal practice.
Some draw a distinction between using ChatGPT in legal practice and using it during law school. In law school, they argue, students need to learn to read and think and write like lawyers, and then only after they master those skills can they choose to take shortcuts in their practice. But using ChatGPT in legal education does not undermine our pedagogical goals. On the contrary, it enhances them. It encourages students to think critically about what lawyers do and about their own potential to contribute to the profession. Prohibiting student use of ChatGPT in the classroom or on assignments ignores the reality that generative AI tools like ChatGPT will become integral to the practice of law. Even worse, ignoring the technology risks further stratifying students into those who use it to their advantage (or at their peril) and those who don’t. Instead, educators should embrace the “ChatGPT revolution” and use it as an opportunity to revisit the ways we teach and assess our students. If a machine can easily do everything we are asking our students to do, then we are not asking enough of our students.
Most of my current students have never diagrammed a sentence. Perhaps my students ten years from now won’t have written a single sentence without the help of AI. We can and should take a moment to lament those losses, but then we must quickly return to the essential goal of our vocation: to equip our students to succeed both today and tomorrow.