CALL member Hannah Rosborough, Instruction & Reference Librarian at the Sir James Dunn Law Library at Dalhousie University in Halifax (Nova Scotia), recently wrote an article on "Thoughts From the Classroom: Addressing Generative AI and Legal Research & Writing". It originally appeared on Slaw.ca on August 16, 2023.

It is republished here with permission of the author, with minor edits.

Generative AI will disrupt legal research. Its negative impact has been highlighted in mainstream media in the UK and the US. Many legal information professionals have valid concerns about how generative AI’s application in legal research may impact the integrity of the profession. Meanwhile, social media (e.g., LinkedIn and Twitter) is flooded with legal tech companies’ commentary on how it can be harnessed to streamline legal research, improving efficiency and productivity. I reached out to several colleagues to hear their thoughts and ideas on how to address this contentious topic in their legal research classrooms.

Determining whether the impact is net negative or net positive will take years, but while we wait, it is important to figure out how to address existing and developing generative AI applications with new legal researchers. Students will use them whether it is discussed in the legal research classroom or not. Once they graduate, students will be faced with AI laws, policies, and guidelines they will be required to follow as it becomes embedded in legal practice and daily-use tools. Ensuring they have the technological literacy to assess generative AI tools is crucial.

For the purpose of this post, I am using definition of generative AI with LLMs as presented by the MLA-CCCC Joint Task Force on Writing and AI, “… computer systems that can produce, or generate, various forms of traditionally human expression, in the form of digital content including language, images, video, and music. LLMs are a subset of generative AI used to deliver text-based formats like prose, poetry, or even programming code.” This post does not address academic integrity explicitly, but there is, of course, overlap given that the context is law school.

There have been multiple accounts of generative AI passing law school exams and the bar, however these are examples of replacing a student or lawyer, not examples of an assistive technology. I intend to teach it explicitly as a supplemental tool in a way that encourages critical thinking and analysis. Most importantly, discussions and activities will communicate that AI should not be used to generate legal research from scratch. Foundational legal research skills are required for productive, accurate use of generative AI. I anticipate it will be an ongoing discussion throughout the year and not isolated to a single class.

I intend to introduce existing general-purpose tools, like ChatGPT, as I teach legal citation to highlight the significance of a well-crafted prompt and why the review and assessment of AI output are crucial. I will continue to do a comparison activity between the common legal research platforms and the variations in search results that I’ve done every year, but with the addition of generative AI legal research tools as a new comparator (e.g. Jurisage). I am also in the early stages of organising a Legal Research Technology day for my LRW class for this upcoming academic year that will inevitably include some generative AI tools, whether standalone or embedded in existing platforms.

To see how other legal research instructors are approaching generative AI, I reached out to several colleagues across Canada and the US to hear their thoughts and get some inspiration. They kindly responded to the following questions:

• 1.      Will you address generative AI in your legal research classroom(s) this upcoming academic year?
• • a.      Have you already introduced it in previous years?
• 2.      Could you briefly explain why you are introducing it to law students?
• 3.     If you are introducing generative AI, could you provide a brief summary of how you intend to introduce it to your students? 

Annette Demers, Reference Librarian at the Paul Martin Law Library, Faculty of Law, University of Windsor

For the fall semester, 2023, I have carved out one week to discuss the use of generative AI for law. I have two guest speakers slated to discuss the issues on the first day of class. On the second day of class that week, it is my intention to provide two very short examples of AI responses to a very contained (and specially designed) legal problem.  One will be from Chat GPT and one will be from a provider that is built specifically for Canadian law.  From there, I’m going to have the students work in groups to analyze the outputs of each. Students will need to submit their analyses and we’ll have an in-class discussion about the students’ findings versus my own observations.

Marcelo Rodriguez, Assistant Librarian & Foreign, Comparative, and International Law Librarian at the Daniel F. Cracchiolo Law Library, University of Arizona

Will you address generative AI in your legal research classroom(s) this upcoming academic year? Or have you already introduced it in previous years? 

Absolutely! I already touched on this topic in the previous academic year. Earlier this year, when ChatGPT made a splash in everyone’s lives, our colleague, Sarah Gotschall made a LibGuide called, ChatGPT and Bing Chat: Generative AI Legal Research. This is a great source which I used to craft a discussion in class around this topic and just assess where the students were, their thoughts and present different angles and perspectives on Generative AI and the Law. When it comes to these emerging and cutting-edge topics, sometimes the students are way more exposed and knowledgeable than you are. Therefore, I strive for a more socratic method and to foster critical thinking. I want to ignite their sense of curiosity, listen to their views and thoughts, and bring it all back to legal research.  

Could you briefly explain why you are introducing it to law students?

There is no doubt that Generative AI is the new kid on the block. However, if you dig a little deeper, you realize that this is yet another head in the longer body of conversations we have had for a while now: Big Data, Artificial Intelligence, Tech Law, Ethics, Bias in AI, etc. I’m convinced that it’s your job as a Legal Research Professor or any professor for that matter, to guide students and help them connect the dots. 

If you are introducing generative AI, could you provide a brief summary of how you intend to introduce it to your students? 

I will probably follow the same approach I did last academic year: discussion, real examples on the big screen and present different points of view regarding the impact of generative AI on the law. This approach is not set in stone and it can’t be. Inevitably, such a new conversation will continue to develop in multiple ways, some of which no one has foreseen at all. In my class, I want that novelty and rapid change to be my students’ catalyst for curiosity. However, as a professor, I strive to connect these fast changing developments and their sense of curiosity to larger conversations which will empower them to be resilient critical thinkers and researchers no matter the new shiny technology that comes along.

Dominique Garingan, Sessional Instructor, University of Calgary Faculty of Law

Will you address generative AI in your legal research classroom(s) this upcoming academic year? Or have you already introduced it in previous years?

I will be addressing generative AI in my advanced legal research (ALR) course this year. I did not address it in my class last year, as it was not as widespread then. Because the course is comprehensive and the class will be meeting once a week, I only have part of one meeting dedicated to the topic. However, I anticipate it coming up in other meetings and being part of ongoing discussions.

Could you briefly explain why you are introducing it to law students?

Generative AI is a subject that taps into both research competence and technological competence. Reflecting on the current legal environment, organizations that students are set to join once they graduate may be introducing or implementing AI policies and practices, conducting trials, and/or evaluating AI’s use cases, risks, and returns on investment in various legal processes or workflows. This may include the use of generative AI in legal research.

Introducing generative AI to students in an ALR course, alongside responsibilities surrounding technological competence, critical thinking, information evaluation, and accountability for errors and omissions, may help students as they enter professional practice, which is arguably becoming more exposed to AI, automation, and other assistive technologies. On a more holistic note, addressing generative AI in ALR may help students allay any personal concerns, detect the limitations and capabilities of the AI tools they will be exploring, determine the immense value they bring as practitioners, and help reframe legal research as a process involving higher-level and critical thinking.

If you are introducing generative AI, could you provide a brief summary of how you intend to introduce it to your students?

I am still thinking about the how and haven’t quite settled on things yet. Because of the evolving nature of AI (the technology) and the course context, I may focus class activities on AI outputs that may, given accountability and extreme caution, be involved in the preparation of research deliverables.

I’m hoping to do a brief lecture on critical evaluation criteria that may be used for evaluating outputs (errors, omissions, and limitations) and an exercise using open-access generative AI tools. For the exercise, I’m hoping to have a question on prompt engineering and another on evaluating gaps and limitations in AI outputs. Because generative AI is a great subject upon which to exercise critical thinking skills, I’m hoping we can do an open discussion or debrief afterward. Much of this will be time-dependent.

Matthew Renaud, Law Librarian, E.K. Williams Law Library, Faculty of Law, University of Manitoba

Will you address generative AI in your legal research classroom(s) this upcoming academic year? Or have you already introduced it in previous years?

Yes, the plan is to address/discuss AI in my Advanced Legal Research course this upcoming academic year. I am aiming to have both a productive discussion on the topic and incorporate it into one of the major student assignments.

Could you briefly explain why you are introducing it to law students?

Law students are already well aware of AI tools like ChatGPT (with many having begun to use them) and we are no longer have the luxury of ignoring this topic. The disruption of the legal profession and legal research by AI is well underway, so not acknowledging it would be doing a disservice to students taking my Advanced Legal Research course.

If you are introducing generative AI, could you provide a brief summary of how you intend to introduce it to your students?

The tentative plan right now is to introduce generative AI in the classroom by highlighting pre-existing tools (such as ChatGPT) to showcase their impact on legal citation and legal research. This will likely take up a single, 3-hour seminar block, but I imagine an organic discussion will take place throughout the semester.

---

The common theme presented is that this is an ongoing, evolving discussion that reiterates the importance of critical thinking and analysis.

Wikipedia was black-listed for academic research in the early aughts as being unreliable and lacking authority. Now there is an understanding that while it cannot replace robust academic resources or steps in the research process, Wikipedia can be used in the same manner as a traditional encyclopedia by providing researchers with an introduction to an unfamiliar topic and identifying key concepts and terms to assist further searches on appropriate platforms to find authoritative sources. It is important to figure out where generative AI may supplement, but not replace, steps in the legal research process. While there are valid reasons for concern, it is important to consider areas of advancement as well.

Addressing generative AI as an assistive tool in the legal research classroom while maintaining focus on teaching foundational legal research skills to new legal researchers should preserve professional integrity and prepare them for a profession undergoing an overdue technological evolution.

Thank you Anette, Marcelo, Dominique, and Matthew for sharing your thoughts and ideas.