Faculty Develop a Student-Centered Approach to AI Guidelines

With the perpetual rise of artificial intelligence, AUC faculty are looking for answers about how students should use AI in the classroom and testing if one set of guidelines could work across disciplines.

At AUC, five faculty members are working with students across nine courses to explore those questions and develop flexible, discipline-specific approaches to using artificial intelligence in teaching and learning. The ongoing research project, “Towards a Guided Application of AI in Teaching and Learning within STEAM Higher Education,” brings together courses spanning construction and mechanical engineering, architecture and Core Curriculum courses.

Led by Mohamed Darwish ’04, ’06, assistant professor of construction engineering, the multidisciplinary approach uses student experiences, attitudes and concerns about AI to help shape guidelines that can evolve along with the technology. 

Maha Bali ’01, co-principal investigator in the research and professor of practice at AUC’s Center for Learning and Teaching, encountered a multitude of questions from fellow faculty regarding the development of AI guidelines in the classroom. 

 "It’s like driving: There are laws you need to follow, but if nobody's watching you, if there are no policemen, what are you going to do?”

“People are always asking for more specific guidelines, and we keep saying, ‘Every discipline and course needs something slightly different,’” Bali said. “This project is trying to see if we can come up with something that would be flexible enough for a diverse group of courses.” 

AI in the Classroom

Rather than developing a single set of rules, the research takes a course-by-course approach. Each participating course begins with a survey, in which students answer questions about their AI perspectives and usage. Instructors then review the findings from their individual courses alongside data from the wider group. With students, the co-PIs create guidelines for AI usage with their students. They then examine the commonalities between the different student-created guidelines.

For Bali, that student involvement is central to the project’s purpose: helping students develop the judgment to make responsible decisions about AI even when no one is there to enforce it.

"Our aim is to have a set of guidelines, but they will ultimately be a first draft for others to build upon: There should be endless numbers of drafts.”

“I hope for my students to develop their own judgment of how they should use AI, whether in my own course or beyond, but they need a bit of scaffolding to do so. It’s like driving: There are laws you need to follow, but if nobody's watching you, if there are no policemen, what are you going to do?” Bali asked. 

For AI, however, there are no clear laws as of yet, Bali stated.

Keeping Pace with AI

As AI further develops, so does the research. The approach is designed to evolve alongside the technology itself. Conducted over four semesters, the research is now in its second year, with one full semester remaining. During that time, new AI tools and capabilities have emerged, and the technology has already advanced in ways that were not present when the project began. The flexibility of the research is a vital element for both Darwish and Bali. 

“It’s a dynamic process. There are issues that are arising and new questions that we weren’t asking before,” Darwish said. Emphasizing student participation, Bali added, “This semester, the plan is to create a new activity and see how students develop AI guidelines for a particular assignment, whereas previously we introduced them to a variety of ethical ways of thinking and asked them to develop generic guidelines based on ethically challenging scenarios.”  

Exploring the effectiveness of these guidelines is not the end of the research. “This is open for change with time, circumstances and norms,” Darwish said. “Our aim is to have a set of guidelines, but they will ultimately be a first draft for others to build upon: There should be endless numbers of drafts.” 

For both Darwish and Bali, the process of co-creating the guidelines with students is important; they hope that the students will process the experience and carry it with them in the future.

By testing AI guidance across different disciplines and incorporating student perspectives, the research addresses a fundamental question: How can educators help students use AI thoughtfully and responsibly in different academic contexts?

More faculty research regarding AI can be found at AUC’s AI information hub.