Custom GPTs present an opportunity for educational innovation. Combining an exploratory and descriptive approach in this emerging arena, we used a case study method to test the potential of a custom GPT in media education. We conducted a qualitative document analysis to assess the textual content generated by a custom GPT and identify themes. To design the custom GPT, we used OpenAI’s GPT-5, released August 7, 2025. Using agentic artificial intelligence (AI) mode, we generated a custom GPT on media ethics. The resulting custom GPT, Media Ethics Instructor (MEI), features an interface designed for college students using a blend of academic and conversational interaction. Our analysis of the content generated by MEI indicates three themes. First, MEI enables Q&A sessions to help students explore key concepts. In responding to questions about core ethical principles, MEI articulated several media ethics concepts, including seeking truth, minimizing harm, and maintaining human oversight when using generative AI. Second, MEI enables simulations with tailored feedback. A simulated student query was posed regarding the use of anonymous sources in journalism. MEI generated a response highlighting the role of unnamed sources in journalism, especially when protecting source confidentiality. Third, MEI distills trends from historical data. MEI revealed that since media reporting of the 1972 Watergate break-in, media usage of anonymous sources has declined steadily. MEI provided its source materials. Limitations of this study include using a single platform (GPT-5) and only one subject (media ethics), and having no assessment of student learning via a custom GPT.

Authors:
John Pavlik, the State University of New Jersey, United States
Tristan Pavlik, Rutgers University, United States

About the Presenter(s)
Professor John Pavlik is a University Professor/Principal Lecturer at Rutgers, the State University of New Jersey in United States

See this presentation on the full schedule – Sunday Schedule

Bookmark