The emergence of generative artificial intelligence (GAI) applications has brought about a significant transformation in postgraduate education. These tools are no longer limited to providing technical support for researchers; rather, they are now capable of generating academic texts, proposing research ideas, suggesting methodological designs, analyzing data, and even drafting research findings. This transformation has created new challenges for academic supervisors, who are increasingly required to reconsider traditional supervision practices, student performance assessment methods, and approaches to monitoring research progress.
One of the most prominent challenges is determining the extent of the student’s genuine research effort in the preparation of a thesis or dissertation. In the past, supervisors could track a researcher’s development by reviewing successive drafts and observing gradual improvements in writing, analytical, and reasoning skills. However, with the use of generative AI, it has become possible to produce coherent academic texts within a very short period, making it difficult for supervisors to distinguish between work genuinely produced by the student and content generated or refined by AI tools.
Another challenge lies in the declining visibility of the traditional indicators used to evaluate a student’s research development. Scientific and linguistic errors, which once revealed areas of weakness and provided opportunities for guidance, are now less apparent. Generative AI often produces polished and well-structured texts that may create a misleading impression of a student’s actual academic competence. Consequently, assessing genuine research capabilities has become more complex than before.
Academic supervision also faces challenges related to verifying the originality of scholarly work. Although plagiarism-detection software is widely used, many AI-generated texts are not classified as plagiarism because they produce linguistically original content, even when the student has not actively contributed to the intellectual construction of the work. As a result, supervisors are confronted with an issue that extends beyond detecting textual similarity to evaluating the originality of ideas, analysis, and conclusions.
Another significant challenge is the absence of clear institutional policies regarding the acceptable use of AI in postgraduate research. The lack of unified guidelines often results in differing interpretations among supervisors concerning the boundary between legitimate AI-assisted support and inappropriate reliance on AI in producing substantive parts of academic work. Such inconsistencies may lead to variations in assessment standards and judgments regarding the quality of students’ research outputs.
Recent literature also highlights ethical and academic accountability challenges. Supervisors are no longer responsible solely for monitoring research procedures and methodological rigor; they are increasingly expected to ensure that students use AI tools transparently and responsibly, and that submitted work reflects genuine understanding of the subject matter. This responsibility becomes particularly important given evidence that AI systems may generate inaccurate information, fabricated references, or unsupported interpretations.
Furthermore, a technological and knowledge gap has emerged between some supervisors and postgraduate students due to the rapid evolution of AI technologies. While many students readily adopt and master new AI applications, some supervisors may lack sufficient training to fully understand their capabilities and limitations. This disparity can affect their ability to guide students effectively and monitor the appropriate use of AI within academic standards.
The growing use of AI has also increased supervisory workloads rather than reducing them in some cases. Supervisors must devote additional time to verifying sources, checking the accuracy of citations, and engaging students in deeper discussions to assess their true understanding of the research topic. Evaluating the final product alone is no longer sufficient; supervisors increasingly need to monitor the entire research process and document the student’s contribution at each stage.
In light of these developments, the central challenge facing academic supervisors is maintaining a balance between leveraging the opportunities offered by generative AI and safeguarding the fundamental objectives of postgraduate education, namely the development of independent researchers who possess critical thinking, analytical skills, and the capacity to produce original knowledge. Consequently, there is a pressing need for universities to establish clear AI policies, strengthen supervisors’ digital competencies, and redesign supervisory practices to meet the demands of the evolving research environment.
