A Systematic Review of the Impact of Generative AI on Postgraduate Research: Opportunities, Challenges, and Ethical Implications

A Systematic Review of the Impact of Generative AI on Postgraduate Research: Opportunities, Challenges, and Ethical Implications | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Systematic Review A Systematic Review of the Impact of Generative AI on Postgraduate Research: Opportunities, Challenges, and Ethical Implications Mabirizi Vicent, Katushabe Calorine, Muhoza B Gloria, Jack Rugasira This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7008752/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The rapid improvement of generative artificial intelligence particularly large language models like ChatGPT, and GPT-4, has presented transformative possibilities in postgraduate research. This systematic review investigates the impact of generative AI tools on the quality, efficiency, ethics, and innovation of postgraduate research. A comprehensive literature search across four databases: Google Scholar, Web of Science, IEEE explore, and Scopus was carried out. Following the PRISMA guidelines, a total of 17 peer reviewed published articles between 2019 and 2025 were selected for a detailed review analysis and these were selected based on relevance to the postgraduate research context, explicit declaration of generative AI application, and reported outcomes. The review identified four thematic areas: (i) research productivity and efficiency, where generative AI has improved academic writing, data analysis, and literature reviews; (ii) cognitive and creative support, where AI helps formulate hypotheses, generate ideas, and refine language; (iii) academic integrity and ethical concerns, highlighting the dangers of plagiarism, data fabrication, and an over-reliance on AI outputs; and (iv) capacity gaps and skills transformation, pointing out the growing demand for postgraduate researchers to receive ethical and AI literacy training. Despite the potential benefits of generative AI tools in democratizing access to research tools and improving productivity among postgraduate researchers, the review discovered that most academic institutions lack robust regulatory frameworks and institutional guidelines. Additional concerns arise from disparities in regional access to cutting-age AI tools, hence compromising the global research equity. The review concludes with a recommendation for a tailored framework for the responsible integration of GenAI into postgraduate research with a focus on institutional oversight, human-AI collaboration, and ethical application. The findings contribute to the current discussion over the future of AI and provide scholars, researchers and policymakers with evidence-based guidance on how to maximize AI’s potential while safeguarding academic integrity. Artificial Intelligence and Machine Learning Special Education Generative AI postgraduate research LLMs research ethics AI integration framework Figures Figure 1 Figure 2 1. Introduction The advent of generative artificial intelligence (GenAI), particularly the large language models (LLMs) led to widespread release and adoption of pre-trained generative transformers such as GPT-3, GPT-4 and Deep Seek within academia (Strzelecki et al., 2024 ). For example Liao et al. ( 2024 ) conducted a large-scale survey involving 816 verified research article authors to explore their perceptions and use of LLMs in research. The study found that 81% of respondents had integrated generative AI tools into various stages of their workflow. Also Jung et al. ( 2024 ) found that 75% of the respondents at the private R1 research university actively use LLMs. These findings highlight the widespread adoption and growing trust in generative AI tools among academic researchers. Large language models are designed to generate human-like prose, code, and even creative content, thereby offering students – especially postgraduate students more options for research assistants (Cheng et al., 2025 ). These models are capable of handling large volumes of data, generate logical writing, condense knowledge and simulate creativity. As a result, they assist postgraduate researchers in all stages of research lifecycle, from proposal development to manuscript writing and dissemination. The integration of GenAI into academic research is rapidly gaining momentum, offering new opportunities to accelerate methodological innovations, speed up literature reviews (Ebadi et al., 2025 ), and improve analytical reasoning (Chandel & Lim, 2024 ; da Silva, 2024 ). Additionally, these tools play a key role in bridging language barriers, hence promoting scholarly communication inclusivity (Chugh et al., 2025 ). Despite these advantages, the integration and adoption of GenAI have got several challenges within postgraduate research. Concerns over academic integrity, diminished critical thinking due to over-reliance on technology specifically AI (Gonsalves, 2024 ), and possible biases in AI-generated information have been widely raised in various academic discussion forums (Cheng et al., 2025 ; Xu et al., 2024 ). These issues raise fundamental questions about the ethical and responsible application of generative AI in postgraduate research settings (Nwozor, 2025 ). Despite the rapid proliferation of GenAI tools in academia, systematic evidence regarding their positive and negative impact remains scattered. As AI technology become increasingly integrated into academia workflows, it is essential to evaluate their influence on postgraduate research. This paper presents a systematic review to critically examine the impact of generative AI on the quality, efficiency, innovation, and ethical standards of postgraduate research practices. 2. Methodology This review integrates the Arksey and O’Malley methodology framework for scoping studies (Arksey & O’malley, 2005 ) with the Preferred Reporting Items for Systematics Studies and Meta-Analyses (PRISMA) guidelines (McGowan et al., 2020 ). While the PRISMA offer a structured approach to enhance the rigor, transparency, and reproducibility of the review process, Arksey and O’Malley’s framework creates a broader aspect of scoping to capture the evolving and interdisciplinary nature of generative AI application in postgraduate research. The methodology followed five steps adopted from both frameworks: (i) identification of the research questions, (ii) identifying relevant studies, (iii) study selection based on inclusion criteria, (iv) charting the data through thematic synthesis, and (v) compiling, summarizing and reporting the results. This hybrid approach ensured both systematic rigor and conceptual coverage, thus, the study incorporated diverse perspectives from empirical studies and position papers within the academic discourse on GenAI. 2.1 Research question identification The study was driven by four main research questions: How has generative AI impacted research productivity and efficiency among postgraduate researchers? In what ways does generative AI support cognitive and creative aspects of postgraduate research? What ethical and academic integrity concerns arise from use of generative AI in postgraduate research? What capacity gaps and skill transformations are emerging as postgraduate researchers integrate generative AI into their workflows? 2.2 Identification of relevant studies A comprehensive literature search was conducted across four databases: Google Scholar, Web of Science, IEEE Explore, and Scopus. A combination of keywords was used, including: “generative AI in research,” “ChatGPT in academia,” “AI and postgraduate education,” and “LLMs in scholarly writing.” The search was restricted to peer-reviewed journal articles and conference papers published between 2019 and 2025. 2.3 Study selection The inclusion criteria required that selected studies: (i) explicitly applied GenAI within the context of postgraduate education; (ii) reported empirical data through case studies or thematic analysis; (iii) were published in English; (iv) appeared in peer-reviewed journals or conference proceedings; and (v) were published between 2019 and 2025. An initial screening of titles and abstracts identified 61 papers. Following a full-text review, a total of 17 studies met the eligibility criteria. 2.4 Charting of the data For each selected study, the following characteristics were extracted: Author(s) and year of publication Study design and methodology AI tool or model investigated Study objective(s) Reported primary outcome on research productivity, cognitive and creative support, ethical considerations, and capacity gaps, and capacity gaps. Key findings The extracted data on the selected characteristics were presented in Table 1 . Author Study design/methodology Study objective Primary outcome Key findings English et al. ( 2025 ) Surveyed 75 PhD students in UK To examine GenAI usage in automating research tasks. Most participants had used GenAI for tasks like idea generation and summarization Increased research efficiency and time for creative tasks. Hoffman, ( 2016 ) Qualitative case study investing 11 female master’s students. To explore GenAI use, benefits, and challenges. GenAI aided in academic writing, planning, and translation. Effectiveness use depended on user engagement and evaluation skills. Usdan & Chang ( 2024 ) Mixed method approach that involved investigating 27 postgraduate students To assess the GenAI’s impact on writing productivity and grades Grades improved from B + to A; writing time reduced by 64.5%. Positive outcomes for both ESL and native English speaker. Noy & Zhang ( 2023 ) Experimental involving 453 professionals. To measure the productivity using ChatGPT in writing tasks. Tasks done 40% faster; 18% improvement in output quality Substantial productivity and quality observed. Chen et al. ( 2025 ) Descriptive study, employing qualitative approach to explore the role of GenAI in group discussions among postgraduate students. To explore the GenAI’s role in group ideation AI simulated ideas in pear review discussions Improved creativity when used to complement human intelligence Puapongsakorn & Brazdeikyte ( 2023 ) A pilot study to investigate how researchers utilize GenAI during ideation processes. To investigate AI in idea generation and sense sensemaking Helped researchers generate diverse perspectives AI added structured frameworks for creativity Moongela et al. ( 2024 ) An experimental study to assess AI-generated prompts in research. To examine GenAI’s impact on divergent thinking Prompt use risked fixation, limiting novelty Critical thinking was emphasized to avoid over reliance on GenAI. Lee et al. ( 2025 ) A survey targeting knowledge workers. To evaluate the impact of GenAI on critical thinking Participants over relied on AI with less mental effort. Risk of cognitive offloading and overconfidence Aytes et al. ( 2025 ) Assessment of theoretical research concepts and proposals. To enhance reasoning ability of GenAI tools. Structured reasoning improved and focus. Balanced AI insights with human research goals observed. Athaluri et al. ( 2023 ) Experimentation of developing 50 research proposals using ChatGPT. To test the citation accuracy in AI-generated content 30% references were had wrong DOI while 16% were fabricated(fake) references Low reliability in references generated by GenAI. Cotton et al. ( 2024 ) Attitude evaluation of students and staff in higher education institution. To assess the attitude and policy gaps in higher education institutes Lack of guidance on acceptable use of AI. Urgent need for clear institutional policies on ethical use of AI Kajiwara & Kawabata ( 2024 ) Pilot study among postgraduate student in access to AI usage digital skills. To promote ethical literacy in AI use. Suggested prompt and citation training. Institutional AI literacy programs were recommended. Zhai et al. ( 2024 ) Competence evaluation of students’ skills in using AI tools. To explore students’ capacity in using AI without close supervision. Unsupervised use linked to reduced critical thinking Guidance is essential to preserve academic rigor. Kaplan ( 2024 ) Assessment of postgraduate program curriculum To assess AI training in curriculum Only 28% included formal GenAI education Recommended revision of postgraduate curriculum to include modules on ethical AI use. Grande et al. ( 2024 ) Observation study involving computing student utilizing AI tools To examine supervisor role in AI usage guidance Supervisors assisted students to ensure ethical use of GenAI tools. Instructor involvement is crucial in AI use. Bianchini et al. ( 2023 ) Assessment of variations in institutional and continental access to AI tools. To assess access disparities in GenAI tools Limited access among certain groups hindered research productivity. Digital divide reduces exposure to GenAI tools. Tadimalla & Maher (2024) Policy proposal to integrate AI in postgraduate education To integrate AI literacy in postgraduate training Advocated ethics modules and AI resource centers Systemic, multi-level integration recommended 3. Results 3.1 Search results Figure 1 presents the search results, screening, exclusion, and final inclusion in the study. 3.2 Discussion of results The analysis revealed four major thematic areas of impact: 3.2.1 Research productivity and efficiency English et al. ( 2025 ) provided valuable insights into the evolving role of GenAI tools in postgraduate research environment, particularly in the United Kingdom. Their survey of 75 doctoral candidates across 19 UK higher education institutions, it was found out that majority of the participants had already used GenAI for example ChatGPT for automating certain research tasks like idea generation, literature search, text summarization and code generation. These applications were perceived as labor-saving, creating more time for researchers to focus on more on analytical and creative tasks (Noy & Zhang, 2023 ). Similarly, Hoffman(2016), conducted a qualitative case study recruiting 11 female masters’ students to explore how they use GenAI tools, their perceived benefit, and challenges they face while utilizing these tools. Through interviews and analysis of ChatGPT conversation logs on the institution network, the study found that GenAI enhanced students’ efficiency in tasks like academic writing, creative project planning, and translation. Despite the GenAI contribution in improving students’ efficiency, the study highlighted that the level of impact is largely dependent on the depth of engagement and user’s critical evaluation skills. In a mixed-methods study with 27 postgraduate students, Usdan & Chang ( 2024 ) investigated the impact of GenAI-assisted writing on productivity and grade in academic assignments. The results of the study showed that both English and ESL students benefited from the use of GenAI tool by improving the grades from average grade B + to grade A and reduced the writing time by 64.5%. Noy and Zhang ( 2023 ) carried out an experimental study to evaluate the productivity of ChatGPT among 453 college-educated professionals tasked with occupation-specific writing assignments. Participants were randomly assigned to two groups i.e. those with and without ChatGPT. Results showed a substantial productivity gain among participants using ChatGPT completing tasks 40% faster than those without ChatGPT. Also, an 18% quality improvement was in the outputs was reported. These studies present a collective demonstration of how GenAI tools play a substantial role in enhancing research productivity and efficiency among postgraduate students and professionals. Key reported achievements include time saving, improved quality of research outputs, and elevated academic performance. These achievements are mainly met where GenAI tools are thoughtfully integrated into research workflow. While these tools provide significant advantages, its positive impact depends on the depth of the user’s engagement, critical thinking, and ethical application. Efficiency gains must therefore be balanced with intentional human oversight to ensure that AI application remains a means to improve the researcher’s intellectual rigor rather than replacing it. 3.2.2 Cognitive and creative support Several studies have documented how GenAI tools been utilized in postgraduate research at different stages to increase productivity and creativity. In this section we discuss some the recent studies that have reported how GenAI has been utilized in research to the cognitive and creative support theme. Chen et al. ( 2025 ) assessed the role of GenAI in conceptual design tasks. In their findings, it was reported that many scholars utilize these tools in the early stages of problem definition and idea generation, thus, enabling them to explore a broader range of concepts and approaches. However, the study raises an importance of human oversight to ensure that the final outcomes align with the study objectives and ethical standards. In the realm of collaborative ideation, Chang & Li ( 2025 ) explored how LLMs are integrated in group brainstorming sessions. According their findings, GenAI tools are capable of improving brain writing produced in peer or group conversation. Study findings shows that AI can be used to stimulate novel ideas, especially where technology is adopted to enhance rather than replace human contribution. Puapongsakorn & Brazdeikyte ( 2023 ) explored the possibility of integrating AI as assistant tools in idea generation, sensemaking, and scientific creativity. Their findings revealed that the integration is possible and provides structured frameworks for developing research ideas hence offering diverse perspectives that may not be possible where traditional methods are solely used. However, the integration of GenAI in cognitive and creative in research workflow has been criticized. Moongela et al. ( 2024 ) explored the impact of AI-generated prompts on design fixation and divergent thinking. Study findings revealed that although GenAI offers valuable inspirations, there is a risk of users becoming overly reliant on AI suggestions, which may limit the novelty of their ideas. Thus, there is a need for critical engagement with outputs produced with the aid GenAI tools to safeguard the autonomy of human creativity. Another study by Lee et al. ( 2025 ) surveyed knowledge workers in a broader context to evaluate the impact of GenAI on critical thinking. Study findings suggest that, while AI tools offer the opportunity for swift information processing, they may also lead to reduced cognitive efforts and overconfidence in AI-generated content. Thus, the study emphasized the importance of nurturing critical thinking alongside AI tool usage. Aytes et al. ( 2025 ) proposed the “Sketch-of-Thought” framework, combining both cognitive inspired reasoning paradigms and linguistic constraints to improve LLM reasoning competence. This move aimed to create a balance between the depths of AI-generated insights with the need for concise and relevant output, thereby empowering researchers in maintaining focus and clarity in their research work. In conclusion, these studies portray a multilayered role of GenAI in augmenting cognitive and creative functions in postgraduate research. Whereas GenAI tools presents a significant role in idea generation and conceptualization, it imperative to approach their use with a critical mindset to ensure that human intelligence and creativity remain central to the research process. 3.2.3 Ethical concerns and academic integrity The increasing integration of GenAI tools into research workflow has prompted significant attention towards emerging ethical concerns and issues of academic integrity. While several issues have been documented, the existing literature consistently highlight three critical concerns associated with AI-assisted writing: AI hallucination, plagiarism, and citation fabrication. In order to generate text, GenAI tools use publicly accessible data, the resultant content may be quite similar to the original source of information (Kacena et al., 2024 ). If this information is not carefully assessed for plagiarism potentials or published without proper acknowledgement of the source can raise concerns of plagiarism and copyright infringement (Cheng et al., 2025 ). In response to user-provided prompts, GenAI tools can produce convincing but entirely factually incorrect information, or illogical and fake, a situation known as AI hallucination (Alkaissi & McFarlane, 2023 ; Buholayka et al., 2023 ). It should be noted that GenAI models are not developed to evaluate authenticity or correctness of the content, instead are trained to predict the next word based on statistical patterns (Cheng et al., 2025 ). Therefore, they are prone to producing fabricated content to fit the user needs, hence resulting in misinformation. Another critical areas where GenAI tools have raised concern is generation of fake references, studies have found that ChatGPT-generated references are mostly fabricated, with only 7% of citations being accurate. AI tools tend to generate output that reproduce or amplifies biases inherent within the source data. For example Athaluri et al. ( 2023 ) used ChatGPT to write 50 research proposal, from results obtained, 38% of the reference in protocols had wrong DOI while 16% were completely fabricated. An investigative study by Cotton et al. ( 2024 ) that evaluated higher education staff and students’ attitude regarding AI-assisted writing raises more concerns. The lack of unified institutional policies and guidelines on what constitutes acceptable use AI is one of the main concerns highlighted in their research findings. The conclusion expressed concerns about urgent need for higher education institutions to develop a context-specific guidelines that defines ethical and responsible use of AI in teaching, learning and research workflows. Birhane ( 2022 ) cautioned against algorithmic biases present in LLMs, which have the potential to strengthen systematic injustices and damaging stereotypes in scholarly discourse. They emphasized that intentional audits and controls are necessary since these biases are often invisible to end users. Furthermore, Kajiwara & Kawabata ( 2024 ) stressed the need for AI literacy in mitigating ethical risks. They suggest that there should mandatory training established in institutions of higher learning focusing on prompt engineering, responsible citation practices, and critical evaluation of the AI-accelerated research outputs. Collectively, these studies reflect a consensus in the in the literature that ethical concerns and academic integrity issues surrounding GenAI in postgraduate research are merely technical issues but are deeply tied to institutional culture, policy and the broader academic ecosystem. As such, addressing these concerns requires multi-layered approach that include ethical education, transparence policy frameworks, and active human oversight. 3.4 Capacity gap and skills transformation Many postgraduate students take AI-generated content at a face value because they lack necessary abilities to evaluate its dependability and quality. Numerous research studies have reported on this allegation. For example Zhai et al. ( 2024 ) discovered that many students use AI in their research workflow without proper guidance, which has been claimed to be a main cause of declined critical thinking among students, potentially weakening academic rigor. Also Kaplan ( 2024 ) found out that only 28% of postgraduate programs include formal instructions on responsible use of AI tools, despite their rapidly growing adoption in academic research. Research has also acknowledged the critical role played by supervisors in promoting ethical use of AI in research. For example Grande et al. ( 2024 ) conducted a study in which computing students discussed LLMs in order to examine their ethical application. Study results demonstrated that instructors were key in assisting students to evaluate content generated with the aid of AI tools and application of AI in moral decision-making. Institutional and regional disparities in access to GenAI tools is another key capacity gap documented responsible for inequalities in research opportunities. For example Bianchini et al. ( 2023 ) reported that researchers in developing or resource-constrained countries often rely on outdated or limited functionality (free packages) of AI tools, if they have access at all. This technology gap not only hinders research productivity, but also limits their exposure to evolving best practices and global academic standards. In order to address these capacity and skills gaps, Tadimalla & Maher (2024) advocate for unified inclusion of AI-related skills in postgraduate programs. Among the integrations suggested, is introduction of ethics modules in research methodology courses, conducting interdisciplinary workshops on responsible use of AI, and creating institutional support systems such as peer mentorship programs or AI resource centers. However, this integration requires coordinated efforts across institutional, pedagogical, and policy level. This capacity building is aimed at boosting individual academic success and promoting inclusiveness in scholarly research while safeguarding academic integrity. 4. Framework for responsible AI integration in postgraduate research Based on the results of the review, we proposed a framework ( see Fig. 2 ) for responsible incorporation of generative AI in graduate studies. The framework is anchored on three interdependent main pillars: (i) institutional oversight, (ii) human-AI collaboration, and (iii) ethical application. 4.1 Institutional oversight The first pillar, institutional oversight, emphasizes the academic institution’s critical role in creating an enabling environment for ethically responsible use of GenAI and adoption. The oversight process is initiated by developing a clear institutional level policies that define boundaries and guidelines on acceptable practices, thereby providing a foundation for ethical conduct and responsible adoption. Capacity building is a core element, this involves developing specialized trainings for both students and supervisors to equip them with required knowledge and skills to navigate the evolving landscape of AI-assisted research workflow and comprehensive curriculum reviews to embed modules for AI literacy, specifically in research methods courses. To ensure continuous support, essential support structures must be established. These include a help desk to offer technical and ethical guidance, and formation of research ethics board equipped with expertise in GenAI technologies. Most importantly these structures are subjected to periodic audits to systematically assess the integration of AI practices across research workflows, hence, enabling the institution to identify gaps and make necessary periodic adjustments. 4.2 Human-AI collaboration Secondly is the human-AI collaboration pillar, which aims at integrating GenAI tools at different levels of research workflow as an enabler rather than substituting human intellectual efforts. The flows suggest key research tasks where AI can be engaged to improve the researcher’s productivity. These tasks include: ideation, searching for literature, summarizing scholarly content, improving readability of written content, grammar and spelling checks, refining content for precision and coherence, and generating structured outline from written materials. However, the framework emphasizes the importance of human cross-validation at every stage to ensure that AI edits reflect the researcher’s voice and critical thinking. It should be noted that while AI tools offer computational assistance, the ultimate responsibility and decision-making remain with the human researcher to safeguard the rigor and creativity that define postgraduate research. 4.3 Ethical use The third pillar, ethical use, is established to promote transparency and accountability across all AI-assisted research processes. Within this tier, the framework recommends the establishment of clear mechanisms to enforce mandatory disclosure of AI-assisted contributions by the researchers. This requirement ensures that researchers openly communicate both the extent and the nature of AI’s role in their work. Besides mandatory disclosure, proper citation practices is emphasized. This is to ensure that AI-generated outputs are appropriately acknowledged and accurately attributed. Combined together, these measures upload scholarly integrity and strengthen ethical standards in research. The ethical commitment is further strengthened by implementing AI detection tools to assist the institution and researchers in identifying instances where AI-generated content is present, hence reducing the risk of undisclosed or irresponsible use. To close the feedback loop, the framework suggests a systematic collection of key performance indicators (KPIs) that monitors ethical compliance, and the impact of AI integration on research quality and outcomes. 5. Conclusion This review has explored the impact of GenAI on postgraduate research through analyzing four thematic areas including research productivity, cognitive and support, ethical concerns, and capacity gaps. The findings highlighted significant potentials of GenAI tools like large language models in improving productivity, supporting ideation, and promoting inclusivity within academic workflows. These tools also offer a seamless literature review process, hence improving the writing quality, and support in conceptual development, hence offering significant value to postgraduate researchers when responsibly applied. The review also identified key challenges associated with academic integrity such as AI hallucination, plagiarism, fabricated citation, and eroding critical thinking resulting from over-reliance. Moreover, low AI literacy, limited institutional guidance and uneven access to GenAI tools, pose significant barriers safeguarding academic integrity amidst of AI adoption. As solution to mitigate these challenges, this paper proposes a three-pillar framework for responsible adoption of GenAI in postgraduate research anchored on institutional oversight, human-AI collaboration, and ethical use. The proposed framework emphasizes the need for integrating AI literacy into teaching curricula, strengthening researcher accountability, and building institutional polices that ensure fairness, transparency, and academic rigor. In conclusion, whereas GenAI offers powerful capabilities to advance postgraduate research, its adoption requires oversight by critical reflection, institutional support, and ethical standards. Future studies should focus on evaluating the evolving impact of GenAI tools and improve strategies for their responsible use, ensuring that technologies advancement aligns with the core values of higher education and scholarly integrity. References Alkaissi, H., & McFarlane, S. I. (2023). Artificial hallucinations in ChatGPT: implications in scientific writing. Cureus , 15 (2). Arksey, H., & O’malley, L. (2005). Scoping studies: towards a methodological framework. International Journal of Social Research Methodology , 8 (1), 19–32. Athaluri, S. A., Manthena, S. V., Kesapragada, V. S. R. K. M., Yarlagadda, V., Dave, T., & Duddumpudi, R. T. S. (2023). Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. Cureus , 15 (4). Aytes, S. A., Baek, J., & Hwang, S. J. (2025). Sketch-of-thought: Efficient llm reasoning with adaptive cognitive-inspired sketching. ArXiv Preprint ArXiv:2503.05179 . Bianchini, S., Müller, M., & Pelletier, P. (2023). Drivers and Barriers of AI Adoption and Use in Scientific Research. ArXiv Preprint ArXiv:2312.09843 . Birhane, A. (2022). The unseen Black faces of AI algorithms . Nature Publishing Group UK London. Buholayka, M., Zouabi, R., & Tadinada, A. (2023). The readiness of ChatGPT to write scientific case reports independently: A comparative evaluation between human and artificial intelligence. Cureus , 15 (5). Chandel, P., & Lim, F. V. (2024). Generative AI and Literacy Development in the Language Classroom: A Systematic Review of Literature. Ubiquitous Learning , 18 (2), 31. Chang, H.-F., & Li, T. (2025). A framework for collaborating a large language model tool in brainstorming for triggering creative thoughts. Thinking Skills and Creativity , 101755. Chen, L., Song, Y., Guo, J., Sun, L., Childs, P., & Yin, Y. (2025). How generative AI supports human in conceptual design. Design Science , 11 , e9. Cheng, A., Calhoun, A., & Reedy, G. (2025). Artificial intelligence-assisted academic writing: recommendations for ethical use. Advances in Simulation , 10 (1), 22. Chugh, R., Turnbull, D., Morshed, A., Sabrina, F., Azad, S., Md Mamunur, R., Kaisar, S., & Subramani, S. (2025). The Promise and Pitfalls: A Literature Review of Generative Artificial Intelligence as a Learning Assistant in ICT Education. Computer Applications in Engineering Education , 33 (2), e70002. Cotton, D. R. E., Cotton, P. A., & Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. Innovations in Education and Teaching International , 61 (2), 228–239. da Silva, A. F. A. (2024). Critical Thinking and Artificial Intelligence in Education . Universidade NOVA de Lisboa (Portugal). Ebadi, S., Nejadghanbar, H., Salman, A. R., & Khosravi, H. (2025). Exploring the Impact of Generative AI on Peer Review: Insights from Journal Reviewers. Journal of Academic Ethics , 1–15. English, R., Nash, R., & Mackenzie, H. (2025). ‘A rather stupid but always available brainstorming partner’: Use and understanding of Generative AI by UK postgraduate researchers. Innovations in Education and Teaching International , 00 (00), 1–15. https://doi.org/10.1080/14703297.2024.2446236 Gonsalves, C. (2024). Generative AI’s Impact on Critical Thinking: Revisiting Bloom’s Taxonomy. Journal of Marketing Education , 02734753241305980. Grande, V., Kiesler, N., & Francisco R, M. A. (2024). Student Perspectives on Using a Large Language Model (LLM) for an Assignment on Professional Ethics. In Proceedings of the 2024 on Innovation and Technology in Computer Science Education V. 1 (pp. 478–484). Hoffman, A. J. (2016). The Impact of Generative AI Tools on Postgraduate Students’ Learning Experiences: New Insights Into Usage Patterns . 32 (3), 255–272. Jung, M., Zhang, A., Lee, J., & Liang, P. P. (2024). Quantitative Insights into Language Model Usage and Trust in Academia: An Empirical Study. ArXiv Preprint ArXiv:2409.09186 . Kacena, M. A., Plotkin, L. I., & Fehrenbacher, J. C. (2024). The use of artificial intelligence in writing scientific review articles. Current Osteoporosis Reports , 22 (1), 115–121. Kajiwara, Y., & Kawabata, K. (2024). AI literacy for ethical use of chatbot: Will students accept AI ethics? Computers and Education: Artificial Intelligence , 6 (March), 100251. https://doi.org/10.1016/j.caeai.2024.100251 Kaplan. (2024). Kaplan Survey: Most Colleges and Universities Take Laissez Faire Approach Towards Use of GenAI in Admissions . https://kaplan.com/about/press-media/caos-survey-colleges-and-universities-approach-towards-gen-ai-admission Lee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems , 1–22. Liao, Z., Antoniak, M., Cheong, I., Cheng, E. Y.-Y., Lee, A.-H., Lo, K., Chang, J. C., & Zhang, A. X. (2024). LLMs as Research Tools: A Large Scale Survey of Researchers’ Usage and Perceptions. ArXiv Preprint ArXiv:2411.05025 . McGowan, J., Straus, S., Moher, D., Langlois, E. V, O’Brien, K. K., Horsley, T., Aldcroft, A., Zarin, W., Garitty, C. M., & Hempel, S. (2020). Reporting scoping reviews-PRISMA ScR extension. J Clin Epidemiol , 123 (177), e9. Moongela, H., Matthee, M., Turpin, M., & van der Merwe, A. (2024). The Effect of Generative Artificial Intelligence on Cognitive Thinking Skills in Higher Education Institutions: A Systematic Literature Review. Southern African Conference for Artificial Intelligence Research , 355–371. Noy, S., & Zhang, W. (2023). Experimental evidence on the productivity effects of generative artificial intelligence. Science , 381 (6654), 187–192. Nwozor, A. (2025). Artificial intelligence (AI) and academic honesty-dishonesty nexus: Trends and preventive measures. AI and Ethics, Academic Integrity and the Future of Quality Assurance in Higher Education , 27 . Puapongsakorn, P., & Brazdeikyte, E. (2023). Exploring the Integration of Artificial Intelligence in the Ideation Stage of Product Development in Swedish Startups: Challenges, Opportunities, and Tool Utilization . Strzelecki, A., Cicha, K., Rizun, M., & Rutecka, P. (2024). Acceptance and use of ChatGPT in the academic community. Education and Information Technologies , 22943–22968. https://doi.org/10.1007/s10639-024-12765-1 Tadimalla, S. Y., & Maher, M. Lou. (2024). AI Literacy for All: Adjustable Interdisciplinary Socio-technical Curriculum. ArXiv Preprint ArXiv:2409.10552 . Usdan, J. H., & Chang, H. (2024). Generative AI ’ s Impact on Graduate Student Writing Productivity and Quality . 1–36. Xu, X., Chen, Y., & Miao, J. (2024). Opportunities, challenges, and future directions of large language models, including ChatGPT in medical education: a systematic scoping review. Journal of Educational Evaluation for Health Professions , 21 . Zhai, C., Wibowo, S., & Li, L. D. (2024). The effects of over-reliance on AI dialogue systems on students’ cognitive abilities: a systematic review. Smart Learning Environments , 11 (1). https://doi.org/10.1186/s40561-024-00316-7 Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7008752","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Systematic Review","associatedPublications":[],"authors":[{"id":478320194,"identity":"04ea7ef6-8641-493d-9f42-af667a8bec88","order_by":0,"name":"Mabirizi Vicent","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA10lEQVRIiWNgGAWjYHACNoYEBgkGNvY2BmYStfAcI0ULGEikEanF4NrhZw8e7rGw55N8lvi5gMFOnkEi/QJ+LbfTzA0Snkkws0mnHZaewZBs2CCRU4BXi+TsBDOJhAMSbGzS6Q3SPAzMQH/lJBDQkv4NpIWHTfJ4828ehnrCWvilc8C2SLBJsB0D2nIYqCX9ACEt5QZALQZsPGlp1jMMjhu28bzBq4MB6IVtD38cqLOXbz9mfLugolqenz39AX49qMAAFE08BqRoAQN2kmwZBaNgFIyC4Q8A9Sg48nBYzZ8AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-8990-4003","institution":"Kabale University","correspondingAuthor":true,"prefix":"","firstName":"Mabirizi","middleName":"","lastName":"Vicent","suffix":""},{"id":478320195,"identity":"c62bc9b2-2c93-4e75-9463-edef86a33837","order_by":1,"name":"Katushabe Calorine","email":"","orcid":"","institution":"Kabale University","correspondingAuthor":false,"prefix":"","firstName":"Katushabe","middleName":"","lastName":"Calorine","suffix":""},{"id":478320196,"identity":"1c1e5f2b-2448-498c-8934-8e81a244c47f","order_by":2,"name":"Muhoza B Gloria","email":"","orcid":"","institution":"Kabale University","correspondingAuthor":false,"prefix":"","firstName":"Muhoza","middleName":"B","lastName":"Gloria","suffix":""},{"id":478320197,"identity":"4d5f397d-c9e5-41bf-8aa4-8a3daf9b65de","order_by":3,"name":"Jack Rugasira","email":"","orcid":"","institution":"Kabale University","correspondingAuthor":false,"prefix":"","firstName":"Jack","middleName":"","lastName":"Rugasira","suffix":""}],"badges":[],"createdAt":"2025-06-30 09:23:24","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":true,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":true},"doi":"10.21203/rs.3.rs-7008752/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7008752/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":85931340,"identity":"3c95e7d0-e539-40f5-ab4c-f9b3411c4757","added_by":"auto","created_at":"2025-07-03 09:25:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":144183,"visible":true,"origin":"","legend":"\u003cp\u003eFlow diagram for the selected studies\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7008752/v1/48c1f6fbdd4a79bcdad82034.png"},{"id":85931341,"identity":"cca97479-8af8-4d2c-a0b1-252b42d90746","added_by":"auto","created_at":"2025-07-03 09:25:17","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":360892,"visible":true,"origin":"","legend":"\u003cp\u003eFramework for integration of GenAI in postgraduate research\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7008752/v1/78324658bf3dc67958863cef.png"},{"id":85931679,"identity":"1c4f9881-cb67-4039-bbbc-bf54d7760295","added_by":"auto","created_at":"2025-07-03 09:33:18","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1214125,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7008752/v1/d2d70015-b7f5-42b6-9d79-01f096059a51.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eA Systematic Review of the Impact of Generative AI on Postgraduate Research: Opportunities, Challenges, and Ethical Implications\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe advent of generative artificial intelligence (GenAI), particularly the large language models (LLMs) led to widespread release and adoption of pre-trained generative transformers such as GPT-3, GPT-4 and Deep Seek within academia (Strzelecki et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For example Liao et al. (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) conducted a large-scale survey involving 816 verified research article authors to explore their perceptions and use of LLMs in research. The study found that 81% of respondents had integrated generative AI tools into various stages of their workflow. Also Jung et al. (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found that 75% of the respondents at the private R1 research university actively use LLMs. These findings highlight the widespread adoption and growing trust in generative AI tools among academic researchers.\u003c/p\u003e \u003cp\u003eLarge language models are designed to generate human-like prose, code, and even creative content, thereby offering students \u0026ndash; especially postgraduate students more options for research assistants (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). These models are capable of handling large volumes of data, generate logical writing, condense knowledge and simulate creativity. As a result, they assist postgraduate researchers in all stages of research lifecycle, from proposal development to manuscript writing and dissemination.\u003c/p\u003e \u003cp\u003eThe integration of GenAI into academic research is rapidly gaining momentum, offering new opportunities to accelerate methodological innovations, speed up literature reviews (Ebadi et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), and improve analytical reasoning (Chandel \u0026amp; Lim, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; da Silva, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Additionally, these tools play a key role in bridging language barriers, hence promoting scholarly communication inclusivity (Chugh et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Despite these advantages, the integration and adoption of GenAI have got several challenges within postgraduate research. Concerns over academic integrity, diminished critical thinking due to over-reliance on technology specifically AI (Gonsalves, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and possible biases in AI-generated information have been widely raised in various academic discussion forums (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Xu et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These issues raise fundamental questions about the ethical and responsible application of generative AI in postgraduate research settings (Nwozor, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eDespite the rapid proliferation of GenAI tools in academia, systematic evidence regarding their positive and negative impact remains scattered. As AI technology become increasingly integrated into academia workflows, it is essential to evaluate their influence on postgraduate research. This paper presents a systematic review to critically examine the impact of generative AI on the quality, efficiency, innovation, and ethical standards of postgraduate research practices.\u003c/p\u003e"},{"header":"2. Methodology","content":"\u003cp\u003eThis review integrates the Arksey and O\u0026rsquo;Malley methodology framework for scoping studies (Arksey \u0026amp; O\u0026rsquo;malley, \u003cspan class=\"CitationRef\"\u003e2005\u003c/span\u003e) with the Preferred Reporting Items for Systematics Studies and Meta-Analyses (PRISMA) guidelines (McGowan et al., \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e). While the PRISMA offer a structured approach to enhance the rigor, transparency, and reproducibility of the review process, Arksey and O\u0026rsquo;Malley\u0026rsquo;s framework creates a broader aspect of scoping to capture the evolving and interdisciplinary nature of generative AI application in postgraduate research.\u003c/p\u003e\n\u003cp\u003eThe methodology followed five steps adopted from both frameworks: (i) identification of the research questions, (ii) identifying relevant studies, (iii) study selection based on inclusion criteria, (iv) charting the data through thematic synthesis, and (v) compiling, summarizing and reporting the results. This hybrid approach ensured both systematic rigor and conceptual coverage, thus, the study incorporated diverse perspectives from empirical studies and position papers within the academic discourse on GenAI.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1 Research question identification\u003c/h2\u003e\n \u003cp\u003eThe study was driven by four main research questions:\u003c/p\u003e\n \u003col style=\"list-style-type: lower-roman;\"\u003e\n \u003cli\u003eHow has generative AI impacted research productivity and efficiency among postgraduate researchers?\u003c/li\u003e\n \u003cli\u003eIn what ways does generative AI support cognitive and creative aspects of postgraduate research?\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWhat ethical and academic integrity concerns arise from use of generative AI in postgraduate research?\u003c/li\u003e\n \u003cli\u003eWhat capacity gaps and skill transformations are emerging as postgraduate researchers integrate generative AI into their workflows?\u003c/li\u003e\n \u003c/ol\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2 Identification of relevant studies\u003c/h2\u003e\n \u003cp\u003eA comprehensive literature search was conducted across four databases: Google Scholar, Web of Science, IEEE Explore, and Scopus. A combination of keywords was used, including: \u0026ldquo;generative AI in research,\u0026rdquo; \u0026ldquo;ChatGPT in academia,\u0026rdquo; \u0026ldquo;AI and postgraduate education,\u0026rdquo; and \u0026ldquo;LLMs in scholarly writing.\u0026rdquo; The search was restricted to peer-reviewed journal articles and conference papers published between 2019 and 2025.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e2.3 Study selection\u003c/h2\u003e\n \u003cp\u003eThe inclusion criteria required that selected studies: (i) explicitly applied GenAI within the context of postgraduate education; (ii) reported empirical data through case studies or thematic analysis; (iii) were published in English; (iv) appeared in peer-reviewed journals or conference proceedings; and (v) were published between 2019 and 2025. An initial screening of titles and abstracts identified 61 papers. Following a full-text review, a total of 17 studies met the eligibility criteria.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e2.4 Charting of the data\u003c/h2\u003e\n \u003cp\u003eFor each selected study, the following characteristics were extracted:\u003c/p\u003e\n \u003cul\u003e\n \u003cli\u003e\n \u003cp\u003eAuthor(s) and year of publication\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eStudy design and methodology\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eAI tool or model investigated\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eStudy objective(s)\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eReported primary outcome on research productivity, cognitive and creative support, ethical considerations, and capacity gaps, and capacity gaps.\u003c/p\u003e\n \u003c/li\u003e\n \u003cli\u003e\n \u003cp\u003eKey findings\u003c/p\u003e\n \u003c/li\u003e\n \u003c/ul\u003e\n \u003cp\u003eThe extracted data on the selected characteristics were presented in \u003cstrong\u003eTable\u0026nbsp;1\u003c/strong\u003e.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\u0026nbsp;\u003ctable id=\"Taba\" border=\"1\" class=\"fr-table-selection-hover\"\u003e\n \u003ccolgroup cols=\"5\"\u003e\u003c/colgroup\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAuthor\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eStudy design/methodology\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eStudy objective\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePrimary outcome\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eKey findings\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEnglish et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSurveyed 75 PhD students in UK\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo examine GenAI usage in automating research tasks.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMost participants had used GenAI for tasks like idea generation and summarization\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIncreased research efficiency and time for creative tasks.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHoffman, (\u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eQualitative case study investing 11 female master\u0026rsquo;s students.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo explore GenAI use, benefits, and challenges.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGenAI aided in academic writing, planning, and translation.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEffectiveness use depended on user engagement and evaluation skills.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUsdan \u0026amp; Chang (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMixed method approach that involved investigating 27 postgraduate students\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo assess the GenAI\u0026rsquo;s impact on writing productivity and grades\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGrades improved from B\u0026thinsp;+\u0026thinsp;to A; writing time reduced by 64.5%.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive outcomes for both ESL and native English speaker.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNoy \u0026amp; Zhang (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eExperimental involving 453 professionals.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo measure the productivity using ChatGPT in writing tasks.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTasks done 40% faster; 18% improvement in output quality\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSubstantial productivity and quality observed.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChen et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDescriptive study, employing qualitative approach to explore the role of GenAI in group discussions among postgraduate students.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo explore the GenAI\u0026rsquo;s role in group ideation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAI simulated ideas in pear review discussions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eImproved creativity when used to complement human intelligence\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePuapongsakorn \u0026amp; Brazdeikyte (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eA pilot study to investigate how researchers utilize GenAI during ideation processes.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo investigate AI in idea generation and sense sensemaking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHelped researchers generate diverse perspectives\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAI added structured frameworks for creativity\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMoongela et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAn experimental study to assess AI-generated prompts in research.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo examine GenAI\u0026rsquo;s impact on divergent thinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePrompt use risked fixation, limiting novelty\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCritical thinking was emphasized to avoid over reliance on GenAI.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLee et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eA survey targeting knowledge workers.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo evaluate the impact of GenAI on critical thinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants over relied on AI with less mental effort.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRisk of cognitive offloading and overconfidence\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAytes et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAssessment of theoretical research concepts and proposals.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo enhance reasoning ability of GenAI tools.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStructured reasoning improved and focus.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBalanced AI insights with human research goals observed.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAthaluri et al. (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eExperimentation of developing 50 research proposals using ChatGPT.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo test the citation accuracy in AI-generated content\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30% references were had wrong DOI while 16% were fabricated(fake) references\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLow reliability in references generated by GenAI.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCotton et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAttitude evaluation of students and staff in higher education institution.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo assess the attitude and policy gaps in higher education institutes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLack of guidance on acceptable use of AI.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUrgent need for clear institutional policies on ethical use of AI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKajiwara \u0026amp; Kawabata (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePilot study among postgraduate student in access to AI usage digital skills.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo promote ethical literacy in AI use.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSuggested prompt and citation training.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInstitutional AI literacy programs were recommended.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eZhai et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCompetence evaluation of students\u0026rsquo; skills in using AI tools.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo explore students\u0026rsquo; capacity in using AI without close supervision.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUnsupervised use linked to reduced critical thinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGuidance is essential to preserve academic rigor.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKaplan (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAssessment of postgraduate program curriculum\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo assess AI training in curriculum\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eOnly 28% included formal GenAI education\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRecommended revision of postgraduate curriculum to include modules on ethical AI use.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGrande et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eObservation study involving computing student utilizing AI tools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo examine supervisor role in AI usage guidance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSupervisors assisted students to ensure ethical use of GenAI tools.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInstructor involvement is crucial in AI use.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBianchini et al. (\u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAssessment of variations in institutional and continental access to AI tools.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo assess access disparities in GenAI tools\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLimited access among certain groups hindered research productivity.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDigital divide reduces exposure to GenAI tools.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTadimalla \u0026amp; Maher (2024)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePolicy proposal to integrate AI in postgraduate education\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTo integrate AI literacy in postgraduate training\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAdvocated ethics modules and AI resource centers\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSystemic, multi-level integration recommended\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Search results\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e presents the search results, screening, exclusion, and final inclusion in the study.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Discussion of results\u003c/h2\u003e \u003cp\u003eThe analysis revealed four major thematic areas of impact:\u003c/p\u003e \u003cdiv id=\"Sec10\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1 Research productivity and efficiency\u003c/h2\u003e \u003cp\u003eEnglish et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) provided valuable insights into the evolving role of GenAI tools in postgraduate research environment, particularly in the United Kingdom. Their survey of 75 doctoral candidates across 19 UK higher education institutions, it was found out that majority of the participants had already used GenAI for example ChatGPT for automating certain research tasks like idea generation, literature search, text summarization and code generation. These applications were perceived as labor-saving, creating more time for researchers to focus on more on analytical and creative tasks (Noy \u0026amp; Zhang, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eSimilarly, Hoffman(2016), conducted a qualitative case study recruiting 11 female masters\u0026rsquo; students to explore how they use GenAI tools, their perceived benefit, and challenges they face while utilizing these tools. Through interviews and analysis of ChatGPT conversation logs on the institution network, the study found that GenAI enhanced students\u0026rsquo; efficiency in tasks like academic writing, creative project planning, and translation. Despite the GenAI contribution in improving students\u0026rsquo; efficiency, the study highlighted that the level of impact is largely dependent on the depth of engagement and user\u0026rsquo;s critical evaluation skills.\u003c/p\u003e \u003cp\u003eIn a mixed-methods study with 27 postgraduate students, Usdan \u0026amp; Chang (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) investigated the impact of GenAI-assisted writing on productivity and grade in academic assignments. The results of the study showed that both English and ESL students benefited from the use of GenAI tool by improving the grades from average grade B\u0026thinsp;+\u0026thinsp;to grade A and reduced the writing time by 64.5%.\u003c/p\u003e \u003cp\u003eNoy and Zhang (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) carried out an experimental study to evaluate the productivity of ChatGPT among 453 college-educated professionals tasked with occupation-specific writing assignments. Participants were randomly assigned to two groups i.e. those with and without ChatGPT. Results showed a substantial productivity gain among participants using ChatGPT completing tasks 40% faster than those without ChatGPT. Also, an 18% quality improvement was in the outputs was reported.\u003c/p\u003e \u003cp\u003eThese studies present a collective demonstration of how GenAI tools play a substantial role in enhancing research productivity and efficiency among postgraduate students and professionals. Key reported achievements include time saving, improved quality of research outputs, and elevated academic performance. These achievements are mainly met where GenAI tools are thoughtfully integrated into research workflow. While these tools provide significant advantages, its positive impact depends on the depth of the user\u0026rsquo;s engagement, critical thinking, and ethical application. Efficiency gains must therefore be balanced with intentional human oversight to ensure that AI application remains a means to improve the researcher\u0026rsquo;s intellectual rigor rather than replacing it.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Cognitive and creative support\u003c/h2\u003e \u003cp\u003eSeveral studies have documented how GenAI tools been utilized in postgraduate research at different stages to increase productivity and creativity. In this section we discuss some the recent studies that have reported how GenAI has been utilized in research to the cognitive and creative support theme.\u003c/p\u003e \u003cp\u003eChen et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) assessed the role of GenAI in conceptual design tasks. In their findings, it was reported that many scholars utilize these tools in the early stages of problem definition and idea generation, thus, enabling them to explore a broader range of concepts and approaches. However, the study raises an importance of human oversight to ensure that the final outcomes align with the study objectives and ethical standards.\u003c/p\u003e \u003cp\u003eIn the realm of collaborative ideation, Chang \u0026amp; Li (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) explored how LLMs are integrated in group brainstorming sessions. According their findings, GenAI tools are capable of improving brain writing produced in peer or group conversation. Study findings shows that AI can be used to stimulate novel ideas, especially where technology is adopted to enhance rather than replace human contribution.\u003c/p\u003e \u003cp\u003ePuapongsakorn \u0026amp; Brazdeikyte (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) explored the possibility of integrating AI as assistant tools in idea generation, sensemaking, and scientific creativity. Their findings revealed that the integration is possible and provides structured frameworks for developing research ideas hence offering diverse perspectives that may not be possible where traditional methods are solely used.\u003c/p\u003e \u003cp\u003eHowever, the integration of GenAI in cognitive and creative in research workflow has been criticized. Moongela et al. (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) explored the impact of AI-generated prompts on design fixation and divergent thinking. Study findings revealed that although GenAI offers valuable inspirations, there is a risk of users becoming overly reliant on AI suggestions, which may limit the novelty of their ideas. Thus, there is a need for critical engagement with outputs produced with the aid GenAI tools to safeguard the autonomy of human creativity.\u003c/p\u003e \u003cp\u003eAnother study by Lee et al. (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) surveyed knowledge workers in a broader context to evaluate the impact of GenAI on critical thinking. Study findings suggest that, while AI tools offer the opportunity for swift information processing, they may also lead to reduced cognitive efforts and overconfidence in AI-generated content. Thus, the study emphasized the importance of nurturing critical thinking alongside AI tool usage.\u003c/p\u003e \u003cp\u003eAytes et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) proposed the \u0026ldquo;Sketch-of-Thought\u0026rdquo; framework, combining both cognitive inspired reasoning paradigms and linguistic constraints to improve LLM reasoning competence. This move aimed to create a balance between the depths of AI-generated insights with the need for concise and relevant output, thereby empowering researchers in maintaining focus and clarity in their research work.\u003c/p\u003e \u003cp\u003eIn conclusion, these studies portray a multilayered role of GenAI in augmenting cognitive and creative functions in postgraduate research. Whereas GenAI tools presents a significant role in idea generation and conceptualization, it imperative to approach their use with a critical mindset to ensure that human intelligence and creativity remain central to the research process.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003e3.2.3 Ethical concerns and academic integrity\u003c/h2\u003e \u003cp\u003eThe increasing integration of GenAI tools into research workflow has prompted significant attention towards emerging ethical concerns and issues of academic integrity. While several issues have been documented, the existing literature consistently highlight three critical concerns associated with AI-assisted writing: AI hallucination, plagiarism, and citation fabrication.\u003c/p\u003e \u003cp\u003eIn order to generate text, GenAI tools use publicly accessible data, the resultant content may be quite similar to the original source of information (Kacena et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). If this information is not carefully assessed for plagiarism potentials or published without proper acknowledgement of the source can raise concerns of plagiarism and copyright infringement (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn response to user-provided prompts, GenAI tools can produce convincing but entirely factually incorrect information, or illogical and fake, a situation known as AI hallucination (Alkaissi \u0026amp; McFarlane, \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Buholayka et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). It should be noted that GenAI models are not developed to evaluate authenticity or correctness of the content, instead are trained to predict the next word based on statistical patterns (Cheng et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Therefore, they are prone to producing fabricated content to fit the user needs, hence resulting in misinformation.\u003c/p\u003e \u003cp\u003eAnother critical areas where GenAI tools have raised concern is generation of fake references, studies have found that ChatGPT-generated references are mostly fabricated, with only 7% of citations being accurate. AI tools tend to generate output that reproduce or amplifies biases inherent within the source data. For example Athaluri et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) used ChatGPT to write 50 research proposal, from results obtained, 38% of the reference in protocols had wrong DOI while 16% were completely fabricated.\u003c/p\u003e \u003cp\u003eAn investigative study by Cotton et al. (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) that evaluated higher education staff and students\u0026rsquo; attitude regarding AI-assisted writing raises more concerns. The lack of unified institutional policies and guidelines on what constitutes acceptable use AI is one of the main concerns highlighted in their research findings. The conclusion expressed concerns about urgent need for higher education institutions to develop a context-specific guidelines that defines ethical and responsible use of AI in teaching, learning and research workflows.\u003c/p\u003e \u003cp\u003eBirhane (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) cautioned against algorithmic biases present in LLMs, which have the potential to strengthen systematic injustices and damaging stereotypes in scholarly discourse. They emphasized that intentional audits and controls are necessary since these biases are often invisible to end users. Furthermore, Kajiwara \u0026amp; Kawabata (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) stressed the need for AI literacy in mitigating ethical risks. They suggest that there should mandatory training established in institutions of higher learning focusing on prompt engineering, responsible citation practices, and critical evaluation of the AI-accelerated research outputs.\u003c/p\u003e \u003cp\u003eCollectively, these studies reflect a consensus in the in the literature that ethical concerns and academic integrity issues surrounding GenAI in postgraduate research are merely technical issues but are deeply tied to institutional culture, policy and the broader academic ecosystem. As such, addressing these concerns requires multi-layered approach that include ethical education, transparence policy frameworks, and active human oversight.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e3.4 Capacity gap and skills transformation\u003c/h2\u003e \u003cp\u003eMany postgraduate students take AI-generated content at a face value because they lack necessary abilities to evaluate its dependability and quality. Numerous research studies have reported on this allegation. For example Zhai et al. (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) discovered that many students use AI in their research workflow without proper guidance, which has been claimed to be a main cause of declined critical thinking among students, potentially weakening academic rigor. Also Kaplan (\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) found out that only 28% of postgraduate programs include formal instructions on responsible use of AI tools, despite their rapidly growing adoption in academic research.\u003c/p\u003e \u003cp\u003eResearch has also acknowledged the critical role played by supervisors in promoting ethical use of AI in research. For example Grande et al. (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) conducted a study in which computing students discussed LLMs in order to examine their ethical application. Study results demonstrated that instructors were key in assisting students to evaluate content generated with the aid of AI tools and application of AI in moral decision-making.\u003c/p\u003e \u003cp\u003eInstitutional and regional disparities in access to GenAI tools is another key capacity gap documented responsible for inequalities in research opportunities. For example Bianchini et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) reported that researchers in developing or resource-constrained countries often rely on outdated or limited functionality (free packages) of AI tools, if they have access at all. This technology gap not only hinders research productivity, but also limits their exposure to evolving best practices and global academic standards.\u003c/p\u003e \u003cp\u003eIn order to address these capacity and skills gaps, Tadimalla \u0026amp; Maher (2024) advocate for unified inclusion of AI-related skills in postgraduate programs. Among the integrations suggested, is introduction of ethics modules in research methodology courses, conducting interdisciplinary workshops on responsible use of AI, and creating institutional support systems such as peer mentorship programs or AI resource centers. However, this integration requires coordinated efforts across institutional, pedagogical, and policy level. This capacity building is aimed at boosting individual academic success and promoting inclusiveness in scholarly research while safeguarding academic integrity.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Framework for responsible AI integration in postgraduate research","content":"\u003cp\u003eBased on the results of the review, we proposed a framework (\u003cb\u003esee\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) for responsible incorporation of generative AI in graduate studies. The framework is anchored on three interdependent main pillars: (i) institutional oversight, (ii) human-AI collaboration, and (iii) ethical application.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Institutional oversight\u003c/h2\u003e \u003cp\u003eThe first pillar, institutional oversight, emphasizes the academic institution\u0026rsquo;s critical role in creating an enabling environment for ethically responsible use of GenAI and adoption. The oversight process is initiated by developing a clear institutional level policies that define boundaries and guidelines on acceptable practices, thereby providing a foundation for ethical conduct and responsible adoption. Capacity building is a core element, this involves developing specialized trainings for both students and supervisors to equip them with required knowledge and skills to navigate the evolving landscape of AI-assisted research workflow and comprehensive curriculum reviews to embed modules for AI literacy, specifically in research methods courses. To ensure continuous support, essential support structures must be established. These include a help desk to offer technical and ethical guidance, and formation of research ethics board equipped with expertise in GenAI technologies. Most importantly these structures are subjected to periodic audits to systematically assess the integration of AI practices across research workflows, hence, enabling the institution to identify gaps and make necessary periodic adjustments.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Human-AI collaboration\u003c/h2\u003e \u003cp\u003eSecondly is the human-AI collaboration pillar, which aims at integrating GenAI tools at different levels of research workflow as an enabler rather than substituting human intellectual efforts. The flows suggest key research tasks where AI can be engaged to improve the researcher\u0026rsquo;s productivity. These tasks include: ideation, searching for literature, summarizing scholarly content, improving readability of written content, grammar and spelling checks, refining content for precision and coherence, and generating structured outline from written materials. However, the framework emphasizes the importance of human cross-validation at every stage to ensure that AI edits reflect the researcher\u0026rsquo;s voice and critical thinking. It should be noted that while AI tools offer computational assistance, the ultimate responsibility and decision-making remain with the human researcher to safeguard the rigor and creativity that define postgraduate research.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Ethical use\u003c/h2\u003e \u003cp\u003eThe third pillar, ethical use, is established to promote transparency and accountability across all AI-assisted research processes. Within this tier, the framework recommends the establishment of clear mechanisms to enforce mandatory disclosure of AI-assisted contributions by the researchers. This requirement ensures that researchers openly communicate both the extent and the nature of AI\u0026rsquo;s role in their work. Besides mandatory disclosure, proper citation practices is emphasized. This is to ensure that AI-generated outputs are appropriately acknowledged and accurately attributed. Combined together, these measures upload scholarly integrity and strengthen ethical standards in research.\u003c/p\u003e \u003cp\u003eThe ethical commitment is further strengthened by implementing AI detection tools to assist the institution and researchers in identifying instances where AI-generated content is present, hence reducing the risk of undisclosed or irresponsible use. To close the feedback loop, the framework suggests a systematic collection of key performance indicators (KPIs) that monitors ethical compliance, and the impact of AI integration on research quality and outcomes.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis review has explored the impact of GenAI on postgraduate research through analyzing four thematic areas including research productivity, cognitive and support, ethical concerns, and capacity gaps. The findings highlighted significant potentials of GenAI tools like large language models in improving productivity, supporting ideation, and promoting inclusivity within academic workflows. These tools also offer a seamless literature review process, hence improving the writing quality, and support in conceptual development, hence offering significant value to postgraduate researchers when responsibly applied.\u003c/p\u003e \u003cp\u003eThe review also identified key challenges associated with academic integrity such as AI hallucination, plagiarism, fabricated citation, and eroding critical thinking resulting from over-reliance. Moreover, low AI literacy, limited institutional guidance and uneven access to GenAI tools, pose significant barriers safeguarding academic integrity amidst of AI adoption.\u003c/p\u003e \u003cp\u003eAs solution to mitigate these challenges, this paper proposes a three-pillar framework for responsible adoption of GenAI in postgraduate research anchored on institutional oversight, human-AI collaboration, and ethical use. The proposed framework emphasizes the need for integrating AI literacy into teaching curricula, strengthening researcher accountability, and building institutional polices that ensure fairness, transparency, and academic rigor.\u003c/p\u003e \u003cp\u003eIn conclusion, whereas GenAI offers powerful capabilities to advance postgraduate research, its adoption requires oversight by critical reflection, institutional support, and ethical standards. Future studies should focus on evaluating the evolving impact of GenAI tools and improve strategies for their responsible use, ensuring that technologies advancement aligns with the core values of higher education and scholarly integrity.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAlkaissi, H., \u0026amp; McFarlane, S. I. (2023). Artificial hallucinations in ChatGPT: implications in scientific writing. \u003cem\u003eCureus\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(2).\u003c/li\u003e\n \u003cli\u003eArksey, H., \u0026amp; O\u0026rsquo;malley, L. (2005). Scoping studies: towards a methodological framework. \u003cem\u003eInternational Journal of Social Research Methodology\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(1), 19\u0026ndash;32.\u003c/li\u003e\n \u003cli\u003eAthaluri, S. A., Manthena, S. V., Kesapragada, V. S. R. K. M., Yarlagadda, V., Dave, T., \u0026amp; Duddumpudi, R. T. S. (2023). Exploring the boundaries of reality: investigating the phenomenon of artificial intelligence hallucination in scientific writing through ChatGPT references. \u003cem\u003eCureus\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4).\u003c/li\u003e\n \u003cli\u003eAytes, S. A., Baek, J., \u0026amp; Hwang, S. J. (2025). Sketch-of-thought: Efficient llm reasoning with adaptive cognitive-inspired sketching. \u003cem\u003eArXiv Preprint ArXiv:2503.05179\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eBianchini, S., M\u0026uuml;ller, M., \u0026amp; Pelletier, P. (2023). Drivers and Barriers of AI Adoption and Use in Scientific Research. \u003cem\u003eArXiv Preprint ArXiv:2312.09843\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eBirhane, A. (2022). \u003cem\u003eThe unseen Black faces of AI algorithms\u003c/em\u003e. Nature Publishing Group UK London.\u003c/li\u003e\n \u003cli\u003eBuholayka, M., Zouabi, R., \u0026amp; Tadinada, A. (2023). The readiness of ChatGPT to write scientific case reports independently: A comparative evaluation between human and artificial intelligence. \u003cem\u003eCureus\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(5).\u003c/li\u003e\n \u003cli\u003eChandel, P., \u0026amp; Lim, F. V. (2024). Generative AI and Literacy Development in the Language Classroom: A Systematic Review of Literature. \u003cem\u003eUbiquitous Learning\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(2), 31.\u003c/li\u003e\n \u003cli\u003eChang, H.-F., \u0026amp; Li, T. (2025). A framework for collaborating a large language model tool in brainstorming for triggering creative thoughts. \u003cem\u003eThinking Skills and Creativity\u003c/em\u003e, 101755.\u003c/li\u003e\n \u003cli\u003eChen, L., Song, Y., Guo, J., Sun, L., Childs, P., \u0026amp; Yin, Y. (2025). How generative AI supports human in conceptual design. \u003cem\u003eDesign Science\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e, e9.\u003c/li\u003e\n \u003cli\u003eCheng, A., Calhoun, A., \u0026amp; Reedy, G. (2025). Artificial intelligence-assisted academic writing: recommendations for ethical use. \u003cem\u003eAdvances in Simulation\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(1), 22.\u003c/li\u003e\n \u003cli\u003eChugh, R., Turnbull, D., Morshed, A., Sabrina, F., Azad, S., Md Mamunur, R., Kaisar, S., \u0026amp; Subramani, S. (2025). The Promise and Pitfalls: A Literature Review of Generative Artificial Intelligence as a Learning Assistant in ICT Education. \u003cem\u003eComputer Applications in Engineering Education\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(2), e70002.\u003c/li\u003e\n \u003cli\u003eCotton, D. R. E., Cotton, P. A., \u0026amp; Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. \u003cem\u003eInnovations in Education and Teaching International\u003c/em\u003e, \u003cem\u003e61\u003c/em\u003e(2), 228\u0026ndash;239.\u003c/li\u003e\n \u003cli\u003eda Silva, A. F. A. (2024). \u003cem\u003eCritical Thinking and Artificial Intelligence in Education\u003c/em\u003e. Universidade NOVA de Lisboa (Portugal).\u003c/li\u003e\n \u003cli\u003eEbadi, S., Nejadghanbar, H., Salman, A. R., \u0026amp; Khosravi, H. (2025). Exploring the Impact of Generative AI on Peer Review: Insights from Journal Reviewers. \u003cem\u003eJournal of Academic Ethics\u003c/em\u003e, 1\u0026ndash;15.\u003c/li\u003e\n \u003cli\u003eEnglish, R., Nash, R., \u0026amp; Mackenzie, H. (2025). \u0026lsquo;A rather stupid but always available brainstorming partner\u0026rsquo;: Use and understanding of Generative AI by UK postgraduate researchers. \u003cem\u003eInnovations in Education and Teaching International\u003c/em\u003e, \u003cem\u003e00\u003c/em\u003e(00), 1\u0026ndash;15. https://doi.org/10.1080/14703297.2024.2446236\u003c/li\u003e\n \u003cli\u003eGonsalves, C. (2024). Generative AI\u0026rsquo;s Impact on Critical Thinking: Revisiting Bloom\u0026rsquo;s Taxonomy. \u003cem\u003eJournal of Marketing Education\u003c/em\u003e, 02734753241305980.\u003c/li\u003e\n \u003cli\u003eGrande, V., Kiesler, N., \u0026amp; Francisco R, M. A. (2024). Student Perspectives on Using a Large Language Model (LLM) for an Assignment on Professional Ethics. In \u003cem\u003eProceedings of the 2024 on Innovation and Technology in Computer Science Education V. 1\u003c/em\u003e (pp. 478\u0026ndash;484).\u003c/li\u003e\n \u003cli\u003eHoffman, A. J. (2016). \u003cem\u003eThe Impact of Generative AI Tools on Postgraduate Students\u0026rsquo; Learning Experiences: New Insights Into Usage Patterns\u003c/em\u003e. \u003cem\u003e32\u003c/em\u003e(3), 255\u0026ndash;272.\u003c/li\u003e\n \u003cli\u003eJung, M., Zhang, A., Lee, J., \u0026amp; Liang, P. P. (2024). Quantitative Insights into Language Model Usage and Trust in Academia: An Empirical Study. \u003cem\u003eArXiv Preprint ArXiv:2409.09186\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eKacena, M. A., Plotkin, L. I., \u0026amp; Fehrenbacher, J. C. (2024). The use of artificial intelligence in writing scientific review articles. \u003cem\u003eCurrent Osteoporosis Reports\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(1), 115\u0026ndash;121.\u003c/li\u003e\n \u003cli\u003eKajiwara, Y., \u0026amp; Kawabata, K. (2024). AI literacy for ethical use of chatbot: Will students accept AI ethics? \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(March), 100251. https://doi.org/10.1016/j.caeai.2024.100251\u003c/li\u003e\n \u003cli\u003eKaplan. (2024). \u003cem\u003eKaplan Survey: Most Colleges and Universities Take Laissez Faire Approach Towards Use of GenAI in Admissions\u003c/em\u003e. https://kaplan.com/about/press-media/caos-survey-colleges-and-universities-approach-towards-gen-ai-admission\u003c/li\u003e\n \u003cli\u003eLee, H.-P., Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., \u0026amp; Wilson, N. (2025). The impact of generative AI on critical thinking: Self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. \u003cem\u003eProceedings of the 2025 CHI Conference on Human Factors in Computing Systems\u003c/em\u003e, 1\u0026ndash;22.\u003c/li\u003e\n \u003cli\u003eLiao, Z., Antoniak, M., Cheong, I., Cheng, E. Y.-Y., Lee, A.-H., Lo, K., Chang, J. C., \u0026amp; Zhang, A. X. (2024). LLMs as Research Tools: A Large Scale Survey of Researchers\u0026rsquo; Usage and Perceptions. \u003cem\u003eArXiv Preprint ArXiv:2411.05025\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eMcGowan, J., Straus, S., Moher, D., Langlois, E. V, O\u0026rsquo;Brien, K. K., Horsley, T., Aldcroft, A., Zarin, W., Garitty, C. M., \u0026amp; Hempel, S. (2020). Reporting scoping reviews-PRISMA ScR extension. \u003cem\u003eJ Clin Epidemiol\u003c/em\u003e, \u003cem\u003e123\u003c/em\u003e(177), e9.\u003c/li\u003e\n \u003cli\u003eMoongela, H., Matthee, M., Turpin, M., \u0026amp; van der Merwe, A. (2024). The Effect of Generative Artificial Intelligence on Cognitive Thinking Skills in Higher Education Institutions: A Systematic Literature Review. \u003cem\u003eSouthern African Conference for Artificial Intelligence Research\u003c/em\u003e, 355\u0026ndash;371.\u003c/li\u003e\n \u003cli\u003eNoy, S., \u0026amp; Zhang, W. (2023). Experimental evidence on the productivity effects of generative artificial intelligence. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e381\u003c/em\u003e(6654), 187\u0026ndash;192.\u003c/li\u003e\n \u003cli\u003eNwozor, A. (2025). Artificial intelligence (AI) and academic honesty-dishonesty nexus: Trends and preventive measures. \u003cem\u003eAI and Ethics, Academic Integrity and the Future of Quality Assurance in Higher Education\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003ePuapongsakorn, P., \u0026amp; Brazdeikyte, E. (2023). \u003cem\u003eExploring the Integration of Artificial Intelligence in the Ideation Stage of Product Development in Swedish Startups: Challenges, Opportunities, and Tool Utilization\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eStrzelecki, A., Cicha, K., Rizun, M., \u0026amp; Rutecka, P. (2024). Acceptance and use of ChatGPT in the academic community. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, 22943\u0026ndash;22968. https://doi.org/10.1007/s10639-024-12765-1\u003c/li\u003e\n \u003cli\u003eTadimalla, S. Y., \u0026amp; Maher, M. Lou. (2024). AI Literacy for All: Adjustable Interdisciplinary Socio-technical Curriculum. \u003cem\u003eArXiv Preprint ArXiv:2409.10552\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eUsdan, J. H., \u0026amp; Chang, H. (2024). \u003cem\u003eGenerative AI \u0026rsquo; s Impact on Graduate Student Writing \u0026nbsp;Productivity and Quality\u0026nbsp;\u003c/em\u003e. 1\u0026ndash;36.\u003c/li\u003e\n \u003cli\u003eXu, X., Chen, Y., \u0026amp; Miao, J. (2024). Opportunities, challenges, and future directions of large language models, including ChatGPT in medical education: a systematic scoping review. \u003cem\u003eJournal of Educational Evaluation for Health Professions\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e.\u003c/li\u003e\n \u003cli\u003eZhai, C., Wibowo, S., \u0026amp; Li, L. D. (2024). The effects of over-reliance on AI dialogue systems on students\u0026rsquo; cognitive abilities: a systematic review. \u003cem\u003eSmart Learning Environments\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(1). https://doi.org/10.1186/s40561-024-00316-7\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Kabale University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Generative AI, postgraduate research, LLMs, research ethics, AI integration framework","lastPublishedDoi":"10.21203/rs.3.rs-7008752/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7008752/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe rapid improvement of generative artificial intelligence particularly large language models like ChatGPT, and GPT-4, has presented transformative possibilities in postgraduate research. This systematic review investigates the impact of generative AI tools on the quality, efficiency, ethics, and innovation of postgraduate research. A comprehensive literature search across four databases: Google Scholar, Web of Science, IEEE explore, and Scopus was carried out. Following the PRISMA guidelines, a total of 17 peer reviewed published articles between 2019 and 2025 were selected for a detailed review analysis and these were selected based on relevance to the postgraduate research context, explicit declaration of generative AI application, and reported outcomes. The review identified four thematic areas: (i) research productivity and efficiency, where generative AI has improved academic writing, data analysis, and literature reviews; (ii) cognitive and creative support, where AI helps formulate hypotheses, generate ideas, and refine language; (iii) academic integrity and ethical concerns, highlighting the dangers of plagiarism, data fabrication, and an over-reliance on AI outputs; and (iv) capacity gaps and skills transformation, pointing out the growing demand for postgraduate researchers to receive ethical and AI literacy training. Despite the potential benefits of generative AI tools in democratizing access to research tools and improving productivity among postgraduate researchers, the review discovered that most academic institutions lack robust regulatory frameworks and institutional guidelines. Additional concerns arise from disparities in regional access to cutting-age AI tools, hence compromising the global research equity. The review concludes with a recommendation for a tailored framework for the responsible integration of GenAI into postgraduate research with a focus on institutional oversight, human-AI collaboration, and ethical application. The findings contribute to the current discussion over the future of AI and provide scholars, researchers and policymakers with evidence-based guidance on how to maximize AI\u0026rsquo;s potential while safeguarding academic integrity.\u003c/p\u003e","manuscriptTitle":"A Systematic Review of the Impact of Generative AI on Postgraduate Research: Opportunities, Challenges, and Ethical Implications","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-03 09:25:13","doi":"10.21203/rs.3.rs-7008752/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"85338f0f-73a9-480b-8090-fd824409136e","owner":[],"postedDate":"July 3rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":50780261,"name":"Artificial Intelligence and Machine Learning"},{"id":50780262,"name":"Special Education"}],"tags":[],"updatedAt":"2025-07-03T09:25:13+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-03 09:25:13","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7008752","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7008752","identity":"rs-7008752","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}