Generative AI as a catalyst for critical thinking: reframing students from consumers to co-creators

Abstract

This study examined students’ AI literacy, perceived learning benefits, and ethical awareness following participation in a structured, AI-integrated classroom activity. The study also investigated the relationships among these constructs and examined whether perceived learning benefits mediate the relationship between AI literacy and ethical awareness. In addition, the study explored whether students’ responses differed according to gender, prior AI experience, and frequency of AI use. The study involved 104 undergraduate students who participated in a hands-on classroom activity over two academic semesters. Students used generative AI tools, including ChatGPT, Google Gemini, and Microsoft Copilot, to prepare and deliver short lessons to their classmates, followed by discussions addressing responsible and ethical AI use. Data were collected using a structured questionnaire measuring AI literacy, perceived learning benefits, ethical awareness, and related aspects of students’ experiences. Overall, students reported positive perceptions of the learning experience. They strongly agreed that AI made learning easier and more enjoyable (M = 4.29), saved time on academic tasks (M = 4.27), and helped improve the quality of their work (M = 4.09). AI literacy was positively correlated with perceived learning benefits (r = 0.559, p  < .001) and ethical awareness (r = 0.386, p  < .001).and no statistically significant gender differences were found across any of the study domains ( p  > .05). The findings suggest that students’ AI literacy is associated with greater perceived learning benefits and ethical awareness. Furthermore, the study examines the potential mediating role of perceived learning benefits in the relationship between AI literacy and ethical awareness. These findings highlight the importance of structured, meaningful classroom activities that encourage students to actively engage with, evaluate, and take responsibility for AI-generated content rather than relying on AI outputs passively.