Policy AI in Education

The advent of artificial intelligence has resulted in a transformative era, causing major changes in many aspects of human life. This evolutionary wave extends its influence to the realm of higher education, potentially reshaping the landscape of academic training and learning methodologies. From past instances of the advent of digital tools, like calculators, computers, mobile phones, and software, it appears that every technology has its influence on the students’ approach. 

For instance, calculators were assessed as useful educational tools that significantly enhance mathematical performance [2]. However, using the calculator was not the only cause. The student’s learning experience also played a significant role. In addition the study shows that the student’s math skills did not necessarily deteriorate due to the use of a calculator [2]. Another study adds that teachers are also an important factor. For example, the way they handle the use of a calculator in class [6]. 

The Internet, which is acknowledged as a foundation for various online learning practices, and its inclusion in course curriculums has a positive influence on students [1]. However, a problem is highlighted in the excessive use of the internet by some students, particularly for non-educational purposes. This poses a potential detriment to their academic goals. To mitigate this issue, implementing a time restriction mechanism on internet usage could prove beneficial. Such a measure would help students maintain a more focused approach towards their educational objectives by limiting distractions stemming from non-academic online activities.

In the realm of computer software utilization, this proactive approach ensures that students develop improved confidence and independence in applying theoretical knowledge in practical contexts [6]. Mobile phones have a worse reputation since they create distractions often causing students to lose focus. However, it also has a positive side: it facilitates planning for students and provides help outside of the school, such as help with homework and communication between students [8]. The diverse impact of different technologies on students’ learning underscores the need for a nuanced understanding, ensuring that their integration enhances rather than hinders the overall educational experience.

Large Language Models (LLMs) represent a recent technological development that has greatly impacted education as well as other aspects of professional life. There are many examples where LLM technology embedded in the curriculum has reshaped the learning approach, with teachers serving as pivotal facilitators for a seamless transition and preparation for this technology. While the integration of LLMs, such as ChatGPT, holds promise for fostering a supportive educational environment, it is imperative to address accompanying risks. For example, plagiarism, students’ reliance on LLM, etc. There are also large amounts of personal data being stored in one place which could lead to leaks [7]. Thus, establishing a policy for the judicious use of such tools in educational settings is essential for the benefit of institutions, students, and educators. 

Considering the potential advantages of employing LLMs as an assistant for instructors and a virtual tutor for students [7], their integration into academic settings is plausible. The automation of diverse skills, such as writing, text creation, and idea generation, prompts a critical examination of educators’ perspectives on automating aspects of the curriculum and its potential impact on learning outcomes and assessments [7]. This dual consideration of both the potentials and risks underscores the importance of a balanced and well-defined approach to the incorporation of LLMs in education. So there are many advantages and disadvantages. This causes questioning among educators, as they do not know how to approach the integration of LLM in teaching. This is where a policy can help because it creates more clarity for educators.

This project endeavors to guide policy development for the incorporation of LLMs into the academic framework of THUAS. It centers around the main question: “What policies are essential to update or modify THUAS curricula for effective LLM integration?”. To address this, two subsidiary questions are posed: “What is the extent of student engagement with premium versus basic LLM features?” and “What are the perspectives of the teaching staff regarding the substitution of certain learning objectives in light of the automation introduced by LLMs?” This study delves into the learning objectives across various faculties and courses within THUAS, scrutinizing the potential impact of LLM integration.

The research highlights that students currently utilize LLMs to assist in completing various academic tasks, including university projects and assignments. However, it is noted that not all faculties and courses at THUAS are adequately equipped for the routine use of AI tools, and some professors may restrict their use due to previously identified risks. Although THUAS has an existing policy concerning the implementation and associated risks of LLMs, it primarily identifies these risks without offering strategies to mitigate them.

This article, therefore, seeks to investigate educators’ attitudes towards the integration of LLMs in educational contexts, aiming to reveal deeper insights. Based on these findings, it proposes comprehensive policies for the judicious use of LLMs by THUAS students and faculty. These policies aim to strike a careful balance between embracing technological innovation and preserving the integrity of traditional educational experiences.


The primary research question driving this study is: “What policies are necessary to ensure that the curricula at The Hague University of Applied Sciences are appropriately updated or modified to facilitate the integration of Large Language Models?”. To provide a comprehensive answer, this research meticulously analyzes data from both surveys and interviews.

Collected data reveals that a significant majority of students, approximately 88%, are utilizing LLMs in their academic pursuits without any formal guidelines. This lack of direction could lead to inefficient or inaccurate use of these tools. Furthermore, it is noteworthy that a 14% of these students have access to premium versions of ChatGPT, potentially giving them an advantage over their peers.

In addition, the analysis of interview data offers in-depth insights into the experiences of faculty members with LLMs within their courses. While most lecturers have not yet adapted their teaching practices to incorporate LLMs, they acknowledge the potential benefits of these tools and are open to exploring their use in education. A recurring theme among the interviewees is the necessity for students to acquire essential skills and knowledge specific to their curricula. This sentiment is encapsulated by one lecturer’s statement: “Our job is to bring out the reflection and make sure students are making active choices.”

Based on this analysis, it becomes evident that several measures need to be implemented to address the concerns of both educators and students regarding the use of LLMs in academic settings.


The research findings and subsequent analysis indicate that LLMs have gained considerable popularity within university settings, establishing themselves as highly beneficial tools. Given this trend, it is likely that LLMs will continue to play a significant role in students’ academic experiences. Consequently, it is imperative for the academic community to adapt to this technological evolution. This adaptation should focus on optimizing the use of LLMs to enhance educational outcomes and align with the evolving landscape of higher education.

To ensure equitable learning opportunities, THUAS is advised to provide all students with uniform resources, specifically the premium version of ChatGPT. In tandem with this provision, it is crucial to offer comprehensive usage guidelines and expert support to both educators and students. This approach will enhance their understanding of the capabilities and limitations of LLMs, like ChatGPT, thereby fostering more efficient and effective use of these tools.

Furthermore, the teaching faculty should re-evaluate their course objectives, ideally in collaboration with an LLM expert. Considering the ability of tools like ChatGPT to facilitate routine written tasks, a feature likely to appeal to students, it is vital to acknowledge the potential for these tools to be misused for time-saving purposes. By redefining learning objectives, educators can ensure that students acquire essential skills and knowledge, thus mitigating the risk of over-reliance on LLMs.

Additionally, while the revision of learning objectives is essential, it remains important to maintain certain core skills and knowledge specific to each course. This implies that complete automation of curricular goals is not advisable. Instead, LLMs should be employed as adjunct, time-efficient tools, similar to the use of calculators in mathematics courses.

Finally, educators must deliberate on appropriate assessment methods aligned with the revised learning goals. This includes deciding whether assessments will permit or even assess the quality of LLM usage by students, or if certain objectives are best achieved without the aid of LLMs. This strategic approach in assessment design will ensure that the integration of LLMs into the academic framework is both effective and aligned with educational objectives.


[1 ] Affum, Mark Quaye, “THE EFFECT OF INTERNET ON STUDENTS STUDIES: A REVIEW” (2022). Library Philosophy and Practice, 6932. https://digitalcommons.unl.edu/libphilprac/6932

[2] Boyle, R. W., & Farreras, I. G. (2015). The effect of calculator use on college students’ mathematical performance. International Journal of Research in Education and Science. https://doi.org/10.21890/ijres.77376 

[3] Chung, L. (2023). What is a good survey response rate for customer surveys in 2022? Delighted. https://delighted.com/blog/average-survey-response-rate

[4] De Winter, J. (2023). Can ChatGPT pass high school exams on English Language Comprehension? ResearchGate. https://www.researchgate.net/publication/366659237_Can_ChatGPT_pass_high_school_exams_on_English_Language_Comprehension

[5] K. E., Monahan, A. B., & Schwarcz, D. (2023). ChatGPT goes to law school. Social Science Research Network. https://doi.org/10.2139/ssrn.4335905

[6] Laborde, C., Kynigos, C., Hollebrands, K., & Sträßer, R. (2006). Teaching and Learning Geometry with Technology. BRILL eBooks (pp. 275–304). https://doi.org/10.1163/9789087901127_011 

[7] Lo, C. K. (2023). What is the impact of CHATGPT on education? A rapid review of the literature. Education Sciences, 13(4), 410. https://doi.org/10.3390/educsci13040410 

[8] Sameer, P. P. (2023). IMPACT OF MOBILE PHONE USAGE ON THE ACADEMIC PERFORMANCE OF STUDENTS? The Online Journal of Distance Education and e-Learning, 11(1). https://www.tojdel.net/journals/tojdel/articles/v11i01c02/v11i01-47.pdf 


Full article is currently under review.

Dr. Hani Al-Ers
Senior Researcher
Hani Al-Ers is a researcher in the field of human-machine interactions. He completed his PhD at the Delft University of Technology at the Interactive Intelligence group of the Faculty Computer Science (EEMCS). Philips Research in Eindhoven sponsored his project which was aimed at improving the user experience of Philips tv sets. He completed 2 post-docs at the Delft University of Technology, during which he managed international consortia on topics such as an improved quality of life for the elderly. Currently, he is conducting research in the field of health and education and he leads the Research Education activities at the Dutch Innovation Factory.