Generative AI Tools for USD Law Faculty

Here are four options for incorporating generative AI into your classroom: All-in, middle ground, exposure, or minimize. 

These strategies were taken from Joseph Regalia, The Case for Iterative Legal Writing Practice with ChatGPT and  USD Center for Educational Excellence and Learning Design Center, Students are Using AI: How to Embrace It; Navigate It; or Minimize It. 

All-in 

• Make sure students are learning the concepts and skills first.
• Allow generative AI use on all submitted and in-class assignments.
• Give guidance on how and when to use at each major state of writing (draft outlines, summarize articles, revise text). 
• Give examples and pitfalls for each stage (guide against using ChatGPT for legal research, preference for tools from Westlaw and Lexis).
• Consider students including reflections on what and how they used it (e.g. Today I learned X about generative AI.  I understood the process behind understanding my prompt to be Y.  This is what led to the output Z.).
• Have students grade a generative AI essay with your rubric.
• Require citations or attribution for generative AI content/assistance. 

Middle ground 

• Identify one major assignment to allow generative AI use that introduces them to the major benefits and weaknesses of GPT. 
  • Use the same strategies for "All-in" for the single assignment. 
• Individually or as a group use generative AI to answer a question in class and paste their responses in a shared google slide deck. 
• Otherwise ban or discourage it for other assignments.

Exposure 

• Use generative AI as a teaching tool in class but do not permit its use on assignments or exams.
• Use a structured generative AI prompt together or individually in-class.
• Information literacy: task students to critique a generative AI created essay or case study using course concepts.
• Help them think through the ethics around using generative AI.
  • Review together the policies and guidelines for ethical use of generative AI.
• Encourage students to use generative AI together or on their own to experiment, practice, and reinforce.

Minimize

• Fool proof your exam/assignments
  • Use in-person exams using pen and paper or ExamSoft.
  • Use oral exams or presentations.
  • Use in-class assignments only (flipped classroom).
• Fine tune essay prompts and assignments 
  • Apply course learning to a very recent news event (post 2021). 
  • Prompts that require course definitions, details from course materials or lecture that are on Blackboard (not the internet), or other knowledge specific to your class.
  • Run your prompt through a generative AI tool to gauge the results.
  • Include elements of self-reflection or personal perspectives.
  • Require research or writing diaries.
• Require discussions about process 
  • Break down assignment into parts with multiple drafts-feedback-revision cycles.
  • Assign reflective journals/logs where students note their efforts, progress, and errors.
  • Have students submit all drafts leading to final elements.
  • Ask process-oriented questions.
  • Use a rubric that includes process elements. 

Tips for assignment creation: 

• Use the “One good assignment” approach to train it on your assignments and then prompt the AI to draft a new assignment.
• Use generative AI to craft a draft for students to evaluate.

Generating a prompt 

• Sample language
  • Use generative AI initially for vocabulary suggestions, before crafting a prompt.
  • Make sure your prompts are specific.
  • Use a prompt generator like AI Mind.
• See our page on prompt engineering
  • Consider training the generative AI tool on a sample answer.
  • Use iterative prompting .

Process-related questions

• Sample language
  • Share the prompt you used to obtain the output you will share.
  • Include the final output from the Generative AI program, as well as any drafts and revised prompts.
  • Using Track Changes (Review Tab - MSWord) or Suggesting (Edit Mode - Google Docs) show your edits and improvements to the original output.
  • Revise the output, personalizing it to your specific choice of program.
  • Disclose the generative AI tool you used and describe how it was used. 
• Example of an assignment with process related questions: Outline for an Online Program Business Plan 

Verification

• Sample language:
  • Review the output from the generative AI tool for accuracy. 
  • Eliminate false cases/laws.
  • Review the analysis of those cases/laws.
• Include this step either on an actual assignment or a canned draft answer to be "checked" by the students.  

• Consider asking students generally about information literacy and how to verify information in an area that you don’t have expertise or experience with.

Reflection/Feedback:

• Sample language: 
  • Critically assess the output from the generative AI program: Think about how you answered the question and how you would “rate” the generative AI tool's answer. If you’re having trouble thinking of how to respond, consider your views in terms of currency, accuracy, authority, relevance, and/or purpose of the answer compared to your answer. 
• Student Feedback on the Use of Generative AI (business plan assignment)