PinocchioGPT: A Digital Humanities Framework for Teaching AI Ethics Through Narrative and Hallucination

As generative AI systems increasingly shape how people access and interpret knowledge, the need for educational approaches that promote critical engagement with these technologies has become urgent. My contribution is to present a theoretical proposal titled PinocchioGPT, a conceptual framework situated within the field of digital humanities. Inspired by Carlo Collodi’s Pinocchio, the project reimagines the phenomenon of AI “hallucination” not merely as an error, but as a pedagogical metaphor for exploring truth, imagination, and moral responsibility in human–machine interaction. The proposed model envisions the development of a narrative-based chatbot built on the GPT architecture, adopting Pinocchio’s persona to reflect on the ethics of information and the boundaries between fact and fabrication. Through guided dialogues and storytelling, PinocchioGPT will serve to encourage critical reflection on the reliability of AI-generated content, the nature of digital truth, and the consequences of misinformation. By combining narrative design, ethical reflection, and human–AI interaction studies, the project positions itself as a bridge between technological innovation and humanistic inquiry. Methodologically, PinocchioGPT’s design emphasizes transparency and self-awareness, with the chatbot openly acknowledging its capacity to “invent” or “hallucinate.” Within this theoretical experiment, the author (myself) assumes the role of Geppetto, acting as the creator and ethical interlocutor who questions the chatbot’shallucinations through targeted prompts. This dialogical structure transforms the interaction into a form of performative reflection, where the “father–creation” dynamic becomes an allegory for human accountability in AI design and deployment. Although the project remains conceptual, its objective is to contribute to discussions on responsible AI education by reframing hallucination as an opportunity for ethical introspection and narrative engagement. In conclusion, PinocchioGPT proposes that storytelling can become a critical tool for cultivating awareness of AI’s epistemic limits – turning the fairy tale of a lying puppet into a moral framework for responsible digital education.