University EFL Teachers’ Perceptions About the Effectiveness of AI-Enhanced E-Assessments in Bangladesh: A Phenomenological Study

Objectives: This study aimed to determine the perceptions of EFL teachers' challenges. In addition, this study investigated potential solutions to mitigate the challenges experienced by EFL teachers. Methods: This research took a qualitative design and, through a purposive sampling method, selected a total of six EFL teachers from three different private universities at a tertiary level. This study collected data through an in-depth interview. Interviews were then coded and analyzed based on thematic analysis in order to identify repeated patterns and insights for a comprehensive exploration of experiences and perspectives among participants. Findings: Most of the EFL teachers in Bangladesh do not consider AI-enhanced e-assessment an effective tool to assess students, and they are skeptical about its efficiency. They prefer pen-and-paper assessment to e-assessment. Infrastructural issues, lack of IT knowledge, lack of training in AI enhanced e-assessment, being unable to prevent e-cheating, plagiarism, and health hazards were some of the major challenges while conducting AI-enhanced e-assessment. Research implications: The study strongly underlines the demand for infrastructure development, rigorous training in using artificial intelligence tools, and effective policy measures to offset e-cheating and plagiarism.