The emergence of generative artificial intelligence (GenAI) has notably transformed the landscape of academic writing, as evidenced by the widespread adoption of tools such as ChatGPT among university students. This technological shift has raised concerns regarding the potential dissemination of misinformation, instances of plagiarism, and an over-dependence on AI. This paper aims to develop a practical framework to help educators effectively guide students in utilizing GenAI for academic writing, integrating both theoretical considerations and empirical field practices. To achieve this, an anonymous online survey was conducted among third-year students at Xi’an Jiaotong-Liverpool University, revealing correlations between the frequency of GenAI use, its impact on academic performance, ethical perceptions, and moral responsibilities. These findings, in conjunction with the author’s firsthand experiences at XJTLU, served as the cornerstone for the development of a robust operational framework that prioritizes inclusion, human agency, AI competency, academic integrity, transparency, and accountability. The primary objective of this framework is to promote ethical GenAI usage among students, thereby cultivating an educational environment that empowers students to engage with AI technology responsibly.

 

Keywords: Academic Writing, Ethical Framework, Responsible AI Usage.

 

 

 

In recent years, generative artificial intelligence (GenAI), which refers to the use of AI technology to create new content such as text, images, and videos (“Generative AI use cases,” n.d.), has significantly affected academic writing. Notable GenAI tools, including ChatGPT, Copilot, Gemini, and Claude, have gained attention for their remarkable content generation capabilities (“Examples of GenAI,” n.d.). OpenAI’s ChatGPT, in particular, has gained significant interest due to its refined language output comparable to human writers (“GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses,” n.d.). According to recent statistics, approximately 86% of university students globally utilize these technologies in their academic pursuits (Ward, 2024). While GenAI integration in academic writing offers potential benefits, such as personalized learning, it has also raised concerns over false information dissemination, plagiarism, academic dishonesty, and overreliance on GenAI (Warschauer et al, 2023; Buriak et al, 2023; Halaweh, 2023; Shidiq,2023). These challenges underscore the need for comprehensive research to facilitate ethical and responsible GenAI use in academic writing.

 

Scholars have extensively examined the advantages and disadvantages of employing GenAI in academic writing. Among its strengths, GenAI helps adjust reading materials to facilitate ideation, generate research outlines, and enhance research performance (Warschauer et al.,2023; Alberth,2023). It is also useful in locating sources, creating research questions, interpreting data output, organizing thoughts, summarizing literature, polishing language, and releasing cognitive resources (Buriak et al., 2023, Alberth, 2023; van Dis et al., 2023; Dergaa et al., 2023; Dwivedi et al., 2023; Nguyen et al., 2024; Enriquez et al., 2024; Kurniati & Fithriani, 2022). However, concerns exist regarding inaccurate data, undermined human agency, and academic misconduct (Miao & Holmes, 2023; Alkaissi & McFarlane, 2023; Meyer et al, 2023). An overreliance on GenAI may hamper critical thinking and other intellectual skills, raising issues related to originality and authenticity in academic assignments (Nguyen et al., 2024; Hyland, 2021; Kasneci et al., 2023; Neumann, 2023). To address ethical considerations, various principles and frameworks focusing on equity, human agency, citizens’ rights, democracy, sustainability, transparency, accountability, security, and safety, have been proposed (Miao & Holmes, 2023; Nguyen et al.,2023). Vetter et al. (2024) also recommend collaborative efforts between teachers and students to establish local ethical guidelines on when and how GenAI can be utilized. However, most of the frameworks available tend to be some general guidelines, a comprehensive and practical framework is needed to guide university students’ ethical use of GenAI in academic writing, particularly from the perspective of teachers.

 

This research aims to establish an operational framework for teachers to guide students’ ethical use of GenAI in academic writing. It investigates existing literature, explores students’ interactions with GenAI in written assignments, investigates their perceptions of these interactions, and examines faculty efforts. The paper is structured as follows: Section 2 introduces the research method, Section 3 presents survey findings, Section 4 discusses identified issues, and Section 5 outlines a proposed framework for ethical GenAI usage. Section 6 provides a conclusion.

 

 

 

Quantitative data were collected from XJTLU students to understand their ethical considerations regarding the use of GenAI. An online questionnaire, adapted from a previously published paper using deontological and teleologicalethical frameworks (Soehardjo et al., 2024), was employed after securing ethical approval from the XJTLU Ethics Committee. Data analysis was conducted using SPSS software. The adapted survey demonstrated high reliability (Cronbach’s alpha coefficient of 0.863) and was interpreted through the lens of deontological ethics. Due to the non-normal distribution of the data, non-parametric tests, including Spearman’s correlation, were employed.

 

A convenience sample of 107 third-year engineering students from XJTLU participated in the survey in the academic year 2024-2025. The majority of participants were male (62.62%), and 75.7% had been using GenAI for over a year.

 

The online survey, presented in English, consisted of 27 items using a five-point Likert scale. The survey assessed participants’ moral obligations, perceptions of benefits and risks associated with GenAI, and their attitudes toward its ethical use in academic writing (see Appendix).

 

 

 

The findings show that 81.31% of participants agree that GenAI enhances work efficiency, and 57% believe it improves overall performance. However, 24.30% express discomfort in using GenAI, and approximately 29% perceive its use as unfair. Less than 47% of respondents believe that detecting GenAI usage is feasible, with 42.05% expressing concerns about potential plagiarism flags. Furthermore, 26.17% of students believe that their peers and instructors would disapprove of their use of GenAI, while 47.66% harbor apprehensions regarding associated risks. Only 14.02% are unsupportive of using GenAI, meanwhile over 82% express confidence in their ability to control its use.

 

Table 1 reveals significant associations between the frequency of GenAI use and students’ ethical perceptions, including feelings of guilt, morality, justification, fairness, acceptance, readiness, and confidence in using GenAI (p<0.05). As students engaged with GenAI more frequently, they exhibited increased confidence and readiness in its utilization and demonstrated a reduced sense of ethical constraint, as evidenced by the linear regression analysis in Table 2.

 

 

 

The analysis presented in Table 3 highlights the critical role of students’ GenAI capabilities and usage in shaping their perceived benefits and moral obligations. Significant Spearman correlation coefficients indicate that higher proficiency in using GenAI is associated with improved task efficiency and academic performance. Additionally, there is a negative correlation between students’ moral obligations and their GenAI capabilities, suggesting that moral obligations decrease as students engage in GenAI use for academic writing. This finding implies that some students may feel less ethically constrained and more inclined to rely on GenAI to complete their assignments.

 

 

Students’ ethical use of GenAI can be influenced by perceived benefits, such as enhanced efficiency and improved academic performance. The negative Spearman correlation coefficients presented in Table 4 indicate a significant relationship between perceived benefits and reduced moral obligations associated with the use of GenAI in academic writing. This finding is consistent with a prior study (Soehardjo et al., 2024) and suggests that as students perceive greater benefits, their sense of moral obligations diminishes. Consequently, students may prioritize these perceived benefits over ethical considerations when using GenAI for academic purposes.

 

 

 

 

The findings presented above shed light on several critical issues that underscore the complexities and ethical considerations surrounding students’ use of GenAI in academic writing.

 

4.1. Growing Disparity among Students

 

The correlation between students’ proficiency in using GenAI and improvements in task efficiency and academic performance raises concerns about a growing divide. Factors such as AI access, technological skills, education, biased training data, trust, privacy, and economic disparity have contributed to a widening digital divide, especially after COVID-19 (Nguyen et al., 2023; Božić, 2023). Nguyen et al. (2024) find that disadvantaged students engage with GenAI tools in a more simplistic and less inefficient manner, potentially leading to unequal academic outcomes and exacerbating existing disparities among students. This highlights the need for further examination of the potential impact on educational equity and inclusivity.

 

4.2. Overconfidence and Risks

 

The observed correlation between students’ confidence in using GenAI and their ethical considerations suggests a potential risk of overconfidence. This overconfidence may lead to the dismissal of GenAI-generated false or biased information, which has been a widespread concern in many existing studies (Warschauer et al., 2023; Alkaissi & McFarlane, 2023; Meyer et al., 2023; Vetter et al., 2024; Mhlanga, 2023). The accuracy and objectivity of GenAI’s contents are determined by its training data, and existing biases and discrimination can be perpetuated and magnified with its algorithms (Mhlanga, 2023). It is crucial to help students obtain a basic understanding of GenAI’s strengths and weaknesses (Warschauer et al., 2023) and guide them to maintain vigilance and remain critical in using it (Vetter et al., 2024).

 

4.3. Academic Integrity and Plagiarism

 

Students’ concerns about GenAI usage and plagiarism highlight complex ethical considerations in academic writing. GenAI-generated text often lacks proper citations or references, potentially leading to unintentional plagiarism (Dergaa et al., 2023). There are apprehensions about the dishonest use of GenAI for cheating, which is difficult to prove due to the lack of accurate detection tools (Halaweh, 2023; Nguyen et al., 2024; Neumann et al., 2023; Rudolph et al., 2023; “Ethical AI for Teaching and Learning,” n.d.). Academic staff struggles to distinguish between AI-generated and human writing, underscoring the need for a deeper exploration of ethical guidelines and preventative measures to maintain academic integrity (Abd-Elaal et al., 2022; Sikander et al., 2023; Chan, 2023).

 

4.4. Impact on Learning and Critical Thinking

 

The observed associations between GenAI usage, perceived benefits, and moral obligations underscore the potential impact on student’s learning experiences and critical thinking skills. Overreliance on GenAI may compromise intellectual skills such as critical thinking, thus posing risks to genuine learning and independent thought (Buriak et al., 2023; Shidiq, 2023; Nguyen et al., 2024; Miao & Holmes, 2023; Kasneci et al., 2023). It is imperative to explore how GenAI technology may influence students’ cognitive development, and ethical reasoning. Strategies must be developed to mitigate any potential negative effects on learning and intellectual growth.

 

 

 

Based on the preceding discourse and the practical measures implemented at XJTLU, a viable framework (see Fig. 1) has been developed to provide ethical guidance to students in the use of GenAI for academic writing. Distinguished by its explicit alignment with the faculty’s viewpoint, empirical field-based methodologies, and a specific emphasis on academic writing, this framework diverges from the one proposed by UNESCO (Miao & Holmes, 2023). It embodies a more pragmatic orientation towards fostering integrity, inclusivity, and the responsible utilization of AI technology within educational contexts.

 

 

Inclusion, as a fundamental component, emphasizes the imperative to ensure equitable access to GenAI and a diversity of information sources. At XJTLU, all students and faculty have had free and universal accessibility to XIPU AI without a time limit provided since 2023. The skills and ethical considerations in utilizing GenAI are being integrated into the workbooks used in classrooms. Additionally, various workshops and seminars are being organized by different departments to ensure that every student is equally included, well-trained, and made aware of biased information, while remaining open to pluralistic opinions. By addressing inclusion, faculty can mitigate potential disparities in access and utilization of GenAI, thereby fostering a more equitable academic environment.

 

Another critical component, human agency, highlights the preservation and enhancement of human cognitive ability and social skills in the context of GenAI utilization. Assignments and assessments are being redesigned at XJTLU to prevent AI usage from impeding students’ cognitive development. For instance, in the module EAP111, both digital and paper-based in-class tests are administered, alongside a written research project report that is divided into several portfolio tasks throughout the academic year. Additionally, speaking coursework requires students to explain and reflect on their research projects and answer related questions. These tasks prioritize critical thinking, creativity, and problem-solving skills over rote memorization and the submission of a singular final product.

 

AI competency serves as a cornerstone in ensuring that students possess a comprehensive understanding of the ethical use of GenAI. Starting in September of 2024, all year-1 students at XJTLU will be required to complete two mandatory modules, namely Essentials of AI and Foundations of AI, aimed at fostering their AI capacity. This initiative also necessitates faculty involvement in continuous professional development, including regular training sessions and participation in community-of-practice activities at multiple levels. These efforts ensure that faculty members are well-equipped to guide students in the ethical use of GenAI in academic writing.

 

Maintaining academic integrity in GenAI-assisted academic writing is increasingly critical as GenAI tools become more integrated into students' and faculty members' academic work. Faculty members must adhere to institutional requirements and policies on academic integrity while enhancing their capacity to detect instances of unethical GenAI use. To prevent the use of fabricated sources, students are required to submit a Source Integration Chart alongside their written assignments, which is rigorously verified by instructors. Clear expectations regarding the ethical use of generative AI tools in their coursework are being communicated to students through explicit AI policies in the syllabus, as well as through embedded discussions during class sessions.

 

Transparency in this framework emphasizes the importance of students’ honest disclosure of GenAI contributions to their academic writing. This involves acknowledging the involvement of GenAI with clear written descriptions of its contributions or screenshots of prompts and AI output. Additionally, students are required to attach an ethics checklist for students to self-review their ethical compliance before submitting their work.

 

Accountability means students must take full responsibility for the accuracy and validity of information generated by GenAI. This encourages students to verify the data produced by AI thoroughly. XJTLU students are taught to critically analyze GenAI output, understand its limitations and potential biases, distinguish between reliable and unreliable sources, and locate original academic sources. By instilling a sense of accountability, faculty can effectively guide students in embracing ethical conduct and scholarly responsibility in their use of GenAI.

 

In summary, the proposed framework, inspired by existing research and informed by first-hand field experience at XJTLU, addresses some of the key challenges associated students’ use of GenAI in academic writing. and is therefore more operational for teachers to effectively navigate the integration of GenAI in academic writing, shaping an educational environment that upholds ethical standards and empowers students to engage with AI technology in a conscientious and scholarly manner.

 

 

 

In conclusion, this study contributes to a deeper understanding of ethical GenAI use in academic writing by identifying critical issues and illuminating the prevalence of GenAI utilization among students. Efforts have been made to develop a practical framework to assist educators in guiding students toward the ethical use of GenAI in academic writing. However, the study’s limitations, such as the use of convenience sampling and reliance on quantitative data, should be acknowledged. Future research should address these limitations by employing diverse and representative sampling methods, as well as incorporating qualitative approaches. The implications of this study extend to the development of educational interventions and policies to promote ethical GenAI usage. These initiatives wouldequip both students and faculty with the necessary knowledge and skills to navigate the ethical complexities of AI technology. Future research in this area should seek to address the identified limitations and further explore the multifaceted nature of ethical GenAI usage in academic writing. This would support the ongoing development of effective ethical frameworks for GenAI integration in academic settings.

 

 

 

Abd-Elaal, E. S., Gamage, S. H. P. W. and Mills, J. E. (2022) Assisting academics to identify computer generated writing. European Journal of Engineering Education, 47(5), 725–745.

 

Alberth. (2023) The use of ChatGPT in academic writing: a blessing or a curse in disguise? TEFLIN Journal, 34(2), 337-352.

 

Alkaissi, H., and McFarlane, S. I. (2023) Artificial hallucinations in ChatGPT: Implications in Scientific Writing. Cureus, 15(2), 1-4.

 

Božić, V. (2023) Artificial Intelligence as the Reason and the Solution of Digital Divide. Language Education & Technology, 3(2), 96-109.

 

Buriak, J. M., Akinwande, D., Artzi, N., Brinker, C. J., Burrows, C., Chan, W. C. W., …Ye, J. (2023) Best Practices for Using AI When Writing Scientific Manuscripts. ACS Nano, 17(5), 4091-4093.

 

Chan, C. K. Y. (2023) A comprehensive AI policy education framework for university teaching and learning. International Journal of Educational Technology in Higher Education 20(38), 1-25.

 

Dergaa, I., Chamari, K., Zmijewski, P. and Ben Saad, H. (2023) From human writing to artificial intelligence generated text: examining the prospects and potential threats of ChatGPT in academic writing. Biol Sport, 40(2):615-622.

 

Dwivedi, Y.K., Kshetri, N., Hughes, L., Slade, E. L., Jeyaraj, A., Kar, A. K., …Wright, R. (2023) Opinion Paper: “So what if ChatGPT wrote it?” Multidisciplinary perspectives on opportunities, challenges and implications of generative conversational AI for research, practice and policy. International Journal of Information Management, 71, 1-63.

Enriquez, G., Gill, V., Campano, G., Flores, T.T., Jones, S., Leander, K. M. … Price-Dennis, D. (2024) Generative AI and composing: an intergenerational conversation among literacy scholars. English Teaching: Practice & Critique, 23(1), 6-22.

 

Ethical AI for Teaching and Learning, (n.d.) Retrieved from: https://teaching.cornell.edu/generative-artificial-intelligence/ethical-ai-teaching-and-learning

 

Examples of GenAI. (n.d.) Retrieved from: https://celt.uky.edu/examples-genai

 

Generative AI use cases. (n.d.) Retrieved from: https://cloud.google.com/use-cases/generative-ai

 

GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses. (n.d.) Retrieved from:

 

Halaweh, M. (2023). ChatGPT in education: Strategies for responsible implementation. Contemporary Educational Technology, 15(2), 1-11.

 

Hyland, K. (2021) Teaching and Researching Writing. 4th edn. England: Routledge.

 

Kasneci, E., Sessler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F. … Kasneci, G. (2023) ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences, 103, 1-12.

 

Kurniati, E. Y. and Fithriani, R. (2022) Post-Graduate Students’ Perceptions of Quillbot Utilization in English Academic Writing Class. Journal of English Language Teaching and Linguistics, 7(3), 437-451.

 

Meyer, J. G., Urbanowics, R. J., Martin, P. C. N., O’Connor, K., Li, R., Peng, P., … Moore, J. H. (2023) ChatGPT and large language models in academia: opportunities and challenges. BioData Mining, 16(20), 1-13.

 

Mhlanga, D. (2023) FinTech and Artificial Intelligence for Sustainable Development. 1st edn. Switzerland: Palgrave Macmillan.

 

Miao, F., Holmes, W. (2023) Guidance for generative AI in education and research. 1st edn. France: UNESCO.

 

Neumann, M., Rauschenberger, M., Schön, M. (2023) “We Need To Talk About ChatGPT”: The Future of AI and Higher Education. 2023 IEEE/ACM 5th International Workshop on Software Engineering Education for the Next Generation (SEENG). Australia: IEEE, 29-32.

 

Nguyen, A., Hong, Y., Dang, B. and Huang, X. (2024) Human-AI collaboration patterns in AI-assisted academic writing. Studies in Higher Education, 49(5), 847-867.

 

Nguyen, A., Ngo, H. N., Hong, Y., Dang, B. and Nguyen, B. T. (2023) Ethical principles for artificial intelligence in education. Education and Information Technologies 28, 4221-4224.

 

Rudolph, J., Tan, S. and Tan, S. (2023) ChatGPT: Bullshit spewer or the end of traditional assessments in higher education. Journal of Applied Learning & Teaching, 6(1), 342-363.

 

Shidiq, M. (2023) The use of artificial intelligence-based ChatGPT and its challenges for the world of education: from the viewpoint of the development of creative writing skills. Proceeding of International Conference on Education, Society and Humanity, Postgraduate program of Nurul Jadid University, Probolinggo, Indonesia, 353-357.

 

Sikander, B., Baker, J. J., Deveci, C. D., Lund, L. and Rosenberg, J. (2023) ChatGPT-4 and Human Researchers Are Equal in Writing Scientific Introduction Sections: A Blinded, Randomized, Non-inferiority Controlled Study. Cureus, 15(11), e49019.

 

Soehardjo, D. W. A.，Hutapea，S. S., Horasi, N., Sentanu, A. G. E., Alnazhary, L. and Larasati, P. D. (2024) Ethical use of artificial intelligence among students: driving factors from deontological and teleological perspectives. 22nd International Conferences on e-Society and 20th International Conference on Mobile Learning, Portugal: IADIS, 101-108.

 

Ward, M. (2024, Sep.10). These 5 AI Tools Are the Most Popular among Students. Retrieved from https://universitybusiness.com/these-5-ai-tools-are-the-most-popular-among-students/

 

Warschauer, M., Tseng, W., Yim, S., Webster, T., Jacob, S.R., Du, Q. and Tate, T. (2023) The Affordances and Contradictions of AI-Generated Text for Second Language Writers. Journal of Second Language Writing, 62, 1-21.

 

van Dis, E.A.M., Bollen, J., Zuidema, W., van Rooij, R. and Bockting, C.L. (2023) ChatGPT: five priorities for research. Nature, 614(7947), 224-226.

 

Vetter, M. A., Lucia, B., Jiang, J. and Othman, M. (2024) Towards a framework for local interrogation of AI ethics: A case study on text generators, academic integrity, and composing with ChatGPT. Computer and Composition, 71, 1-12.

 

 

 

Questionnaire used

 

   I agree to participate in the survey study

a. Yes (Please proceed)   b. No (This is the end)

 

Section 1 Demographic questions

 

    My gender is _____________.

 

    I have used GenAI for：  

a. More than one year   b. between six and twelve months  c. less than six months

 

    How often do I use GenAI  

a. at least once a day      b. at least once a week

c. at least once a month  d. less than once a month

 

 

Section 2 Survey questions

1. I feel guilty if I used GenAI to finish my assignment
2. Using GenAI for the benefit of finishing my assignment is in contradiction with my principle.
3. In general, the use of GenAI for task or study is eligible.
4. I feel it unfair to use GenAI to finish the assignment.
5. I feel wrong when using GenAI to finish assignment.
6. I feel inequitable when using GenAI to finish assignment.
7. When I used GenAI, my grading increases.
8. When I used GenAI, my task will be faster.
9. When I used GenAI, my performance improves.
10. When I used GenAI, it was possible to find out.
11. When I used GenAI, that was possibly detected as plagiarism.
12. When I used GenAI, I got a reduced assessment.
13. When I used GenAI, most of my friends and lecturers would disagree.
14. Most of my friends and lecturers would look down on me if I used GenAI.
15. None of my friends and lecturers would be acceptive when I used GenAI.
16. My friends think that the behavior of using GenAI is wrong.
17. I think that the behavior of using GenAI is not wise.
18. I think that the behavior of using GenAI was dangerous.
19. I think that the behavior of using GenAI was a bad idea.
20. In general, I will not support the use of AI for assignment.
21. I easily used GenAI for my assignment.
22. I know how to use GenAI for my assignment.
23. I am capable of using GenAI for my assignment.
24. Using GenAI to do my assignments is within my control.
25. I do not use GenAI to finish my assignment.
26. I do not have a plan for using GenAI to finish my assignment.
27. I will not recommend using GenAI to finish an assignment.

(Adapted from Soehardjo, Hutapea, Horasi, Sentanu, Alnazhary & Larasati, 2024)

 

In-class Pictures