){kind=logo}
){kind=icon}
){kind=logo}
){kind=icon}
Learning to use ChatGPT is already part of university students’ usual practices to support content comprehension and exam preparation. As a fifth-semester student in the bachelor’s degree in Strategy and Business Transformation (LAET) program at Tecnologico de Monterrey, Laguna Campus, I have observed that students commonly use Artificial Intelligence (AI) tools in their day-to-day lives. This is supported by 18 interviews I conducted with my classmates at the university and documented information. Among the most relevant comments my colleagues shared with me was that interacting with an assistant who answers their questions, offers examples, and serves as a “study partner” motivates them to prepare. Listening to students from different semesters report using ChatGPT as an academic tutor during their individual study sessions led us to propose the approach outlined in this article. The other author, a Business School professor at Tec de Monterrey, believes it is essential to integrate the student perspective into this proposal for innovative experiences that transform education. In this article, we share student and teacher perspectives on our experience with the design, implementation, and preliminary evaluation of an intelligent tutor to support individual study and exam preparation, along with the results obtained.
Currently, the debate on the use of AI in education has shifted towards the need to integrate it responsibly, harnessing its potential to improve learning experiences. To face this challenge, our central objective in this proposal was to incorporate the needs and visions of students and teachers into a tool that improves the teaching and learning experience. Academically, we seek to contribute to the current conversation on the ethical and pedagogical use of AI in higher education, emphasizing its potential to support individual study and assessment preparation rather than to substitute for the learning process.
How do students use AI?
Classroom experience shows that AI-based tools are increasingly common among university students to support their learning. This trend has been confirmed by various studies indicating that AI is used to manage information, summarize content, clarify questions, prepare for exams, and support academic tasks such as writing and paraphrasing (Rahman et al., 2025). In general, students perceive these tools as assistants that facilitate the understanding of complex topics and optimize their study time (Rodríguez-López, 2024).
In Mexico, studies report broad acceptance and largely positive perceptions of AI applications for academic purposes. Several authors document that students recognize its usefulness in personalizing learning and improving content comprehension, although relevant concerns have also been identified in the academic field. (Chávez & De los Ríos, 2025; Aguilar-Guggembul et al., 2025; Rodríguez-López, 2024). These include risks to academic integrity, the instrumental use of AI for repetitive tasks, and a possible reliance on these tools to the detriment of critical thinking (Abe & Kazumu, 2025; Ríos-Zabala, 2025).
Many concerns about the use of AI in education have centered on the dilemma of whether students use these tools to work for themselves. So, how do you get these tools to work with them? In this scenario, there is a need to design specific strategies that guide the use of AI toward ethical and formative practices in individual study and exam preparation, where chatbots can serve as allies in learning rather than substitutes for cognitive effort.
Let’s get to work
From our experiences using generative AI, we identified its potential as a personal tutor available to students 24 hours a day. Such an assistant can answer specific questions, adapt to each student’s knowledge level, and offer examples aligned with their interests. With this idea in mind, we decided to develop our own chatbot to help students understand the topics covered in class and simulate final exam questions.
One of the first questions when starting this project was: What AI tool can we use to host our chatbot? Several factors led us to choose ChatGPT. First, the research consulted shows that it is the most widely used generative AI application among students, so they are familiar with it. Second, it incorporates the GPTs application (Generative Pretrained Transformer) to create personalized chatbots, and its use is relatively simple. In addition, this app includes a wizard that makes it easy to create prompts that meet the required specifications. Third, we believe a personalized chatbot is the best option for a tailored experience aligned with our class’s teaching policies and specific objectives.
The following question framed our second big uncertainty: What characteristics should our personalized GPT have? Initially, we thought it should be a tutor who would explain the topics covered in class to the students. Several arguments emerged, both for and against the idea. In favor, students would have a tutor available who could explain the topics using the same terms and concepts covered in class. In this way, a virtual teacher who could attend to each student’s specific needs would be available at any time and in any place. However, copyright and student privacy needed to be considered. To address these elements, we decided to make it a rule for the chatbot never to ask for personal data beyond the name. In addition, it was considered pertinent to feed the tool only with the class presentations, never with the original sources.
On the other hand, we found that we were overlooking one of the ways students use AI most in the study process: preparing for exams. Thus, we decided to include in the personalized GPT a bank of questions that simulate those of the exam (not the same ones) so that the student could practice and identify the topics that require more attention.
Main functions of the custom GPT for the business class
Our first custom GPT pilot had three key functions. First, Tutor Mode, which explains the topic step-by-step, uses clear, real-life examples and connects the concepts to those covered in the course. Second, Practice Mode helps reinforce learning through multiple-choice questions on specific class topics and provides students with immediate feedback on both mistakes and correct answers. Finally, Exam Mode simulates short exams from a pre-loaded question bank, grades answers, and provides feedback.
We decided to conduct the first pilot with a group of 25 first-semester students at Tec de Monterrey’s Laguna Campus Business School. These students had the opportunity to interact with the AI-based academic tutor for a week before the final exam. According to information collected by the GPT, during the test period, the chatbot had more than 100 conversations and was rated an average of 4.7 out of 5 by its users.
Additionally, as part of the pilot evaluation, a questionnaire based on the Technology Acceptance Model (TAM) (Davis et al., 1989) was administered. Four elements were measured to determine user satisfaction: 1) perceived usefulness and enjoyment; 2) ease of use; 3) attitude towards use; and 4) intention to use the system.
Results
Among the most important results, 59% of the sample described the chatbot as very useful for their final exam preparation. No respondents felt the instrument was useless. In addition, 68% of users found the tool very easy to use.
Regarding attitude towards use, 68% of the sample reported that the GPT tutor helped them a lot in understanding the class topics. This result is consistent with how students used the tool: the most popular function was the Tutor, followed by Practice Mode and Exam Mode.
Interestingly, 77% of the sample members stated that the tool clearly helped them identify their weakest areas of knowledge. In addition, 81% of participants reported increased motivation to study, suggesting a positive impact on their learning habits. This is reflected in the fact that most of them reported using the tool at least three times.
Finally, 91% of the sample indicated that they would recommend the GPT tutor to their peers as study support. Among the positive aspects highlighted were the tool’s versatility, the ability to summarize and review previous topics, access to immediate feedback, ease of use, and a comprehensive design that allows users to find explanations and practice exams in the same space.
However, it wasn’t all smooth sailing; we also identified areas for improvement. For example, the sample students noted that some topics were not explained properly by the tool. Upon our review, we realized that the information we had provided to the chatbot was very brief, giving us our first lesson: an AI-based chatbot tutor is only as smart as the information it receives. Therefore, make sure to feed it with plenty of information on all the topics you want it to address.
Another aspect to consider is that users without the paid version of ChatGPT have a limited number of interactions, which was one of the main limitations identified during the pilot. To address this situation, we designed the GPT to provide more concise, structured responses and to group frequent requests into a single interaction, reducing the need for multiple user queries.
With this first implementation, the AI-based academic tutor ceased to be a conceptual idea and became a concrete classroom experience. The pilot enabled observation of how students interacted with the tool, which aspects they valued most, the challenges they encountered in daily use, and how they integrated the intelligent tutor into their study habits. These initial findings laid the foundation for reflecting on the real scope of solutions like this and extracting relevant lessons to improve future implementations across various contexts.
Reflection
The experience of designing and implementing an AI-based academic tutor made it clear that the debate should not focus on whether students use these tools, but on how they use them and for what purpose. As mentioned throughout this article, AI is already a natural part of students’ study habits, and, if well designed, it can become a powerful ally for better understanding content, practicing autonomously, and studying with greater motivation. It is not a question of replacing the teacher; instead, the chatbot serves as a permanent companion that extends the classroom beyond class hours and responds to a real need of contemporary students: to learn at their own pace and receive immediate feedback.
At the same time, this experience places a clear responsibility on institutions and teachers. Integrating AI into the educational process is not only a technological decision, but also a pedagogical and ethical one. Preliminary results show that when used for understanding, practice, and reflection, not just for answers, AI can strengthen critical thinking and foster better study habits. The challenge in the future will be to continue learning to design intentional tools, building practical limits, and leveraging their full potential to transform how we teach and learn in higher education.
Integrating students’ perspectives into the design of educational proposals is both valuable and necessary. Listening to them and giving their views a real place in these processes recognizes that their experience has a unique and transformative value. With the student body’s vision, we enrich initiatives with ideas that would not be visible from other perspectives. Precisely, the development of this pilot demonstrated the value of working together: teachers and students learning from each other and sharing perspectives, which enabled the development of a richer, more humane, and more complete proposal. Between the teaching experience and the student voice, something truly significant emerged: a solution that not only addresses a need but also reflects what is possible to create through authentic collaboration. This shared project is a clear example of what can be achieved when students are no longer seen solely as recipients but as co-creators of meaningful educational solutions.
Finally, if you are a teacher with questions about using AI in the classroom or are interested in exploring how these tools can support your students’ learning, we are happy to talk. We will be happy to share our experience, exchange ideas, and accompany you in designing initiatives tailored to your context.
About the Authors
Ivis Gutiérrez Guerra (i.gutierrez@tec.mx) holds a PhD in Economics and is a full-time professor at Tecnológico de Monterrey Business School on the Laguna Campus.
Ares Daniel Rodríguez Castillo (A01230990@tec.mx) is a student of Strategy and Business Transformation at Tecnológico de Monterrey, with experience in consulting and business analysis projects. He is interested in innovation, data-driven decision-making, and working in entrepreneurial ecosystems.
References
Aguilar-Guggembul, J., Ventura-Mojica, E., Ávila-García, L., Validivia-Villanueva, G. N. (2025). Use of Artificial Intelligence Tools at a Technological University in the East of the State of Mexico. International Journal of Professional Business Review, 10 (7), 01–09. https://doi.org/10.26668/businessreview/2025.v10i7.5575.
Chávez Márquez, I.L., De los Ríos Chávez, H. J. (2024). Personal and academic use of artificial intelligence in university students: exploratory study. Apertura, 12 (1), 54–69. http://doi.org/10.32870/Ap.v17n1.2604.
Chidinma, A. E., & Yinka, K. R. (2025). Undergraduate students’ perception of artificial intelligence for enhancing educational practices: Opportunities, challenges, and ethical considerations. Global Journal of Social Sciences Studies, 11(1), 30–41. https://doi.org/10.55284/gjss.v11i1.1411
Davis, F., Bagozzi, R. & Warsaw, P. (1989). User acceptance of computer technology: a comparison of two theoretical models.Management sciences, 35(8), 983-1003. https://doi.org/10.1287/mnsc.35.8.982
Kamalov, F., Santandreu Calonge, D., Gurrib, I. (2023). New Era of Artificial Intelligence in Education: Towards a Sustainable Multifaceted Revolution. Sustainability 2023, 15 (12451), 1–27. https://doi.org/10.3390/su151612451.
Levy, D. & Pérez Albertos, Á. (2025). Teaching Effectively with ChatGPT (2nd ed.) (Published on Amazon Books).
Rahman, M.M., Chung, C.C., Shanana, A., Chu Hua, S.A.S., Thiviyashree, Elleyajasmin (2025). Utilization of Generative Artificial Intelligence Technologies as Learning Tools among University Students: A Cross-Sectional Study. International Journal of Online and Biomedical Engineering (iJOE), 21(7), pp. 4–14. https://doi.org/10.3991/ijoe.v21i07.54227.
Ravšelj, D., Keržič, D., Tomaževič, N., Umek, L., Brezovar, N., Iahad, N., et al.(2025). Higher education students’ perceptions of ChatGPT: A global study of early reactions. PLoS ONE, 20(2), 1-53. https://doi.org/10.1371/journal.pone.0315011
Ríos Zavala, A. (2025). Perceptions and Uses of Artificial Intelligence in Students of the Bachelor’s Degree in Preschool Education at ByCENED. RedCA Journal, 8 (15), 8–15. https://revistaredca.uaemex.mx/article/view/27111
Rodríguez López, M. (2025). Perceptions and uses of artificial intelligence in students of Biotechnologist Chemistry and Materials Science Engineering at the UJED. RedCA Magazine, 8 (15), 65-69. https://revistaredca.uaemex.mx/article/view/27123
Editing
Edited by Rubí Román (rubi.roman@tec.mx) – Editor of the Edu bits articles and producer of The Observatory webinars- “Learning that inspires” – Observatory of the Institute for the Future of Education at Tec de Monterrey.
Translation
Daniel Wetta
){kind=icon}
){kind=icon}
This article from Observatory of the Institute for the Future of Education may be shared under the terms of the license CC BY-NC-SA 4.0 
