Rebooting the Writing Classroom: Jeanne Beatrix Law on Custom GPTs, Student Agency, and Ethical AI
How a Kennesaw State professor and her students are transforming writing instruction with custom generative AI tools and a focus on responsible collaboration.
Dr. Jeanne Beatrix Law is a professor of English at Kennesaw State University whose research and teaching center on generative AI, writing studies, and ethical prompt engineering. In this interview, Law discusses how AI is being used in Undergraduate and Graduate Writing Courses.
Source note: This is an edited interview adapted from a narrated video submitted to OpenAI.
Intro
Dr. Jeanne Beatrix Law is pioneering new territory in writing instruction at Kennesaw State University. As both a researcher and educator, she’s been at the forefront of integrating generative AI into undergraduate and graduate writing courses. Her approach? Treat AI not as a shortcut, but as a thought partner—one that students can collaborate with, question, and learn from, all while maintaining their own voice and agency. In this conversation, Dr. Law shares how she and her students have developed custom GPTs to engage with primary sources and open educational resources, the frameworks that ensure ethical and effective use, and the real-world impact of these tools in the classroom.
The Interview
Q: Dr. Law, you’ve been experimenting with generative AI in writing courses since ChatGPT first became widely available. What led you to focus on custom GPTs as a teaching tool?
Law: The big draw for me was the ability to create truly interactive, student-centered learning experiences. Custom GPTs let us move beyond generic AI responses and build tools that are grounded in our own course materials and ethical frameworks. For example, my students and I have developed bots that engage directly with primary research sources—like oral history interviews from the Atlanta Student Movement—and others that act as instructional assistants based on open-source writing textbooks. The goal is always to cultivate engagement with the writing process, not just the product.
Q: Can you walk us through how you ensure these AI tools actually support student learning, rather than just making things easier or more automated?
Law: Absolutely. The foundation is what I call the Rhetorical Prompting Method. It’s a framework for human-AI collaboration that puts learner agency first. We teach students to use precise language in their prompts, to reflect on the outputs, and—crucially—to always revise, so that the human voice and intent remain central. It’s really a riff on best practices in process writing, but adapted for the AI era. We also developed the Ethical Wheel of Prompting, a set of questions students ask themselves during the process to ensure their intent guides the collaboration. These frameworks are open educational resources, and they’re based on research with thousands of students.
Q: Let’s talk about your Atlanta Student Movement GPT. How did you and your students build it, and what does it do?
Law: We started with the Atlanta Student Movement Oral History Project, which is a rich archive of interviews, transcripts, and documents from the 1960 sit-ins in Atlanta. We transformed this archive into a custom GPT by uploading only these primary sources into the knowledge base—no outside data, no web search. Then we trained the model to ensure historical accuracy and respectful representation. The bot is programmed to help students analyze key figures, documents, and rhetorical strategies, and even to suggest thesis directions. But it’s also designed with guardrails: it won’t write papers for students, and it always reminds them to review and revise any output. When a student says, “I’m finished,” the bot prompts them to double-check their work and provides the original source for transparency.
Q: That’s a thoughtful design. How accessible is this process for other educators who might not have a technical background?
Law: That’s one of the most exciting parts—no coding is required. Everything is done through clear written instructions, both for the bot and for the students. We provide OER guides on how to write precise instructions for custom GPTs, and the knowledge base is just a matter of uploading documents—PDFs or Word files. You can choose to enable or disable web search, depending on your goals. It’s a very approachable way for educators to create a custom repository tailored to their curriculum.
Q: You also mentioned an AI assistant for first-year writing, built on the OpenStax writing guide. How does that work, and what’s the impact on students?
Law: The process is similar. We uploaded the OpenStax writing guide—an open educational resource—chapter by chapter into the GPT’s knowledge base. The bot is trained to help students with specific writing issues, like building a bibliography, by pointing them to the relevant sections in the guide. Again, it’s not connected to the internet, so students stay within the curated knowledge base. This encourages them to iterate and engage deeply, rather than just looking for quick answers. And, as with the other bot, it’s programmed never to write the paper for the student. The feedback from students has been really positive—they appreciate having a safe, focused space to experiment with generative AI and improve their writing process.
Q: What do you see as the broader implications of this work—for writing instruction, for student agency, and for the future of AI in education?
Law: I think we’re at a moment where we can choose to use AI to empower students, not replace them. When we design tools that foreground process, reflection, and ethical collaboration, we help students develop not just writing skills, but critical thinking and digital literacy. These custom GPTs are just one example of how generative AI can be harnessed to support, rather than undermine, the core values of education. My hope is that more educators will experiment with these approaches and adapt them for their own contexts.
What Stands Out
Core idea: Dr. Law’s work centers on using custom GPTs as interactive, student-centered tools that support—not shortcut—the writing process. By anchoring AI in primary sources and open educational resources, she ensures that technology enhances, rather than replaces, authentic learning.
Classroom design: Her classrooms are built around frameworks like the Rhetorical Prompting Method and the Ethical Wheel of Prompting, which teach students to collaborate with AI thoughtfully and responsibly. The custom GPTs are easy to build—no coding required—and are carefully programmed to keep students engaged in iterative, reflective writing.
Student impact: Students gain hands-on experience with both primary research and generative AI, learning to ask better questions, analyze sources, and maintain their own voice. The bots provide guidance, structure, and ethical guardrails, helping students develop confidence and agency in their writing.
Transferable lesson: Dr. Law’s approach demonstrates that with the right frameworks and resources, any educator can harness AI to foster deeper engagement, critical thinking, and ethical collaboration—skills that are essential far beyond the writing classroom. Take a look at one example here.
Bio
Dr. Jeanne Beatrix Law is a professor of English at Kennesaw State University whose research and teaching center on generative AI, writing studies, and ethical prompt engineering. She developed the Rhetorical Prompting Method and the Ethical Wheel of Prompting, and coordinates the university’s graduate certificate in AI and writing technologies. She researches how writers engage with generative AI models by conducting Think Aloud Protocols, a research method where writers and their screens/voices are recorded as they complete writing tasks with AI. She has presented research at Cambridge and Oxford. She has published in multiple academic and public venues. Her AI frameworks are open access and license CC:BY.