Afghan lecturers’ awareness, application, and concerns on ChatGPT integration in higher education

Abstract

The rapid expansion of generative artificial intelligence has prompted higher education systems worldwide to reassess the pedagogical, ethical, and institutional implications of ChatGPT. However, empirical evidence from conflict-affected and under-resourced educational settings remains limited. This mixed-methods study investigated Afghan lecturers’ awareness, usage patterns, perceived effectiveness, and concerns regarding ChatGPT at Kandahar University. Quantitative data were collected from 98 lecturers using a validated survey instrument, complemented by semi-structured interviews with 10 purposively selected participants. Findings revealed a notably low and uneven adoption of ChatGPT, with most lecturers engaging with the tool only rarely or not at all. Respondents acknowledged its usefulness for research drafting, translation, and academic writing, concerns regarding plagiarism, student overreliance, inaccurate outputs, and fabricated references significantly undermined acceptance. Correlation analysis indicated that perceived effectiveness positively predicted usage, whereas concerns exerted a suppressive effect. Qualitative insights further contextualized these trends, showing that institutional ambiguity, limited digital readiness, and ethical uncertainty weakened perceived usefulness and ease of use. The study extends the Technology Acceptance Model (TAM) by demonstrating that contextual and ethical risk perceptions can override cognitive acceptance determinants in low-resource systems. Practical implications include the urgent need for institutional policies, targeted professional development, and AI literacy initiatives to ensure responsible integration.

Ähnliche Arbeiten