Generative AI (GenAI) tools invite us to take a step back and examine our pedagogy with fresh eyes. Even as the capabilities of these tools expand and change, the principles of good, evidence-informed pedagogy continue to apply.
We realize everyone is in a different place on their journey of learning what GenAI tools can do and considering whether or how to integrate them into their teaching. Wherever you are, we encourage you to be intentional and reflective about your goals, to communicate your expectations around GenAI use clearly to students, and to avoid attempting to police perceived student misuse. Increasingly, AI is integrated into the programs and apps students use to write and edit. Moreover, AI detection tools have been shown to be deeply flawed, flagging human-written text as AI-generated and vice-versa.
If you have never experimented with a GenAI tool, we recommend trying some out. Different tools have different strengths. Consider trying Microsoft Copilot (GW has an enterprise license), ChatGPT, Claude, or Google Gemini for generating texts, and Adobe Firefly or DALL-E for generating images. Ask some of these tools to complete assignments you currently use in your classes to see what they can and cannot do well. Do not input any sensitive data, including student work (which may be a FERPA violation).
After reviewing the tips below, consider joining us for a workshop. The Center for Teaching Excellence regularly holds workshops that address teaching with GenAI. You can see this semester’s full schedule on the LAI Events page. We also can offer a custom workshop for your department or other unit. Please email us at teaching@gwu.edu to discuss.
Examine your objectives and purpose
Where you choose to integrate GenAI tools into your classroom will depend on your discipline, your course, and your course’s place in a curriculum. Considering your course objectives can help you clarify where you want students to use GenAI tools and where you want them to do their own intellectual work. Where does the learning happen, and where can GenAI tools help students get to where the learning happens?
Consider the skills and knowledge your students already have and those they will need outside of your classroom. How can you convey the value of those things to your students to motivate them? Do your students have the necessary expertise yet to judge the output of a GenAI tool?
Have a syllabus policy
After considering your course objectives, write a clear policy for your syllabus outlining what GenAI uses are allowed or disallowed and when. Different assignments may have different policies, so remind students of your policy frequently. GW faculty member Ryan Watkins has developed a free tool that creates a chart laying out what GenAI uses are permitted for an assignment.
Discuss GenAI with your students
You do not need to be an expert on GenAI to discuss it with your students. Consider the following questions:
- How is your field approaching GenAI? What are professionals using it for?
- What do scholars in your field have to say about the many ethical issues raised by GenAI tools? Are they concerned with the environmental impact, the human labor required during training of the tools, intellectual property concerns, the reproduction of biases, data privacy issues, or concerns about equity (more powerful tools cost money)?
- What questions do your students have about GenAI? How are your students using GenAI?
In addition, consider exploring this Student Guide to Artificial Intelligence from Elon University and sharing and discussing it with your students. The guide includes practical tips, addresses student FAQs, and discusses how students' careers may intersect with GenAI.
Design assignments purposefully and transparently
Many GW faculty have adopted the research-backed Transparency Framework for their assignments. The framework involves describing:
- Purpose: Why are students doing this work?
- Task: What do students need to do?
- Criteria: How will students know they have done the work well?
Including these three components in your assignment handouts increases student motivation and success. Consider building in information about GenAI to each component:
- Purpose: Why should students use or avoid using GenAI in a particular way? How will following your guidelines for appropriate use support their success in your class, their major, or a future career?
- Task: What steps can GenAI support? What steps must be done without support from GenAI?
- Criteria: Will students need to cite or reflect on their GenAI use? Will they be graded on the quality of their final product or will you also evaluate some aspect of their process, including ways in which they interacted with GenAI?
Value the process
Learning requires attention to the process as well as the final product, but students often focus on the product. Think about how your course can help students value the process and figure out if they are on the right track before they reach the final paper or big exam. You might break larger projects into smaller pieces (some of which may benefit from using GenAI tools), or provide more opportunities for students to reflect about their own processes. You might ask students questions like:
- What kinds of feedback would help you move this project forward?
- What do you feel proud of in this draft?
- Did you find GenAI helpful when completing this task?
- What did GenAI get wrong and how did you evaluate the quality of the GenAI output?
Consider the value of the human
The way we bring our unique human traits to GenAI can create the most productive learning encounters. As our Georgetown colleague J Palmeri puts it, “If we ask students to start a task by consulting AI, they may defer too strongly to the AI’s advice. If we engage students in human writing and speaking activities before bringing in AI, we can then frame AI as an additional conversation partner whose ideas may or may not be helpful.”
You can also consider your students’ work processes, needs, and anxieties. Students are most likely to use GenAI tools in an unauthorized manner when they are working on a large task at the last minute. Breaking down large assignments into smaller, scaffolded tasks can help students budget their time and feel less overwhelmed. Your late work policy can also help students seek support from you or other authorized resources rather than turning to GenAI.
Explore GenAI tools
As you explore different GenAI tools, consider trying to use them in the following ways:
- Enter an assignment prompt you give to students to see what different GenAI tools generate. ChatGPT, CoPilot, and Claude produce very different results, so it is worth trying more than one. Consider asking the tool that produces the strongest result to refine the response further, thinking about how your students might interact with the tool.
- Sample prompt: “I am a [type of student, class, etc.: e.g., first-year college student in introductory biology] writing [type of assignment: e.g., a lab report on cell division]. [Paste in the assignment prompt, including any sections or subsections you require.] Can you complete it for me? Go step by step.”
- Give the tool an assignment prompt that you use with students and ask it to help you make it more transparent or identify questions that students might have. What seems clear to us may not be readily obvious to our students, and AI can help us bridge that gap.
- Sample prompt: “I am a college professor teaching [discipline, type of class: e.g., a senior capstone seminar]. I give my students the following assignment prompt [include the prompt]. Using the transparent assignment design framework, can you identify what questions students might have and suggest changes that I could make to increase the assignment’s transparency?”
- Ask a tool to identify the steps a student will need to take to complete an assignment successfully. When you think about the “tasks” step of the transparent assignment design framework discussed above, it is sometimes hard to identify all the sub-steps you are asking from your students. Having a clearer sense of the sub-steps can help you plan how to support students where needed.
- Sample prompt: “I am trying to scaffold an assignment for students. They are undergraduates in an introductory history course writing a 12–15-page term paper. I would like to think of all the tasks my students must complete. Be as detailed as possible: so, if one step is ‘researching,’ what tasks go into that?”
- Try your own prompt! Prompts that include a role for the GenAI tool (e.g., student, researcher, concerned citizen), specific context (e.g., in a specific class, about to go to a research site, writing an op-ed), and an expected output (e.g., number of ideas, length of response, in a specific style, at a specific reading level) often give more useful responses. You can ask for it to go step by step, or ask for revisions or an entirely new output.
