Ethical challenges and opportunities in ChatGPT integration for education: insights from emerging economy

Abstract

This study delves into the ethical implications of implementing ChatGPT in Cambodian educational settings chosen for their unique pedagogical challenges in integrating AI tools. The research explores the use of ChatGPT as a supplementary educational tool to create study materials, facilitate discussions, and provide student feedback. Data from 297 students and teachers in various Cambodian educational institutions were collected through structured questionnaires and analyzed using partial least squares structural equation Modeling (PLS-SEM). The study systematically investigates the influence of data privacy concerns, perceived bias, fairness, and teacher-student dynamics on the perceived ethicality and subsequent adoption of ChatGPT. The results show that concerns about data privacy and perceived bias significantly and negatively impact ethical perceptions. However, fairness is also a mediating factor in mitigating these adverse effects. For instance, when AI tools provide equitable support, concerns about bias tend to diminish, thereby improving ethical perception. Furthermore, the reduction in face-to-face interactions, including personalized guidance, spontaneous discussions, and non-verbal cues, negatively affects the perceived ethicality of AI tools by undermining trust and reducing meaningful human connections. These insights provide practical recommendations for educational institutions to ensure responsible and equitable integration of AI technologies, ultimately supporting an ethically sound and effective learning environment.