This project investigated the integration of Artificial Intelligence and its impact on the education scene, and how Indonesian university students perceive the role of AI in their education. We discovered that:
1. Unequal access to internet infrastructure can magnify educational inequality.
Despite the overwhelming positive reviews and perceptions of using AI from student participants in the survey, we should keep in mind the potential implications of AI serving as a magnifier of unequal accessibility in education, as well as other negative consequences, such as unequal access to internet infrastructure and gaps in AI literacy across provinces. Regional differences in the responses suggest that access to AI tools is uneven across Indonesian provinces. For example, provinces with stronger internet access and greater technological access appear too have a stronger institutional integration of technology in coursework, as well as an intent to continue to use AI in education. On the other hand, regions with less resources may have access to less reliable and compatible AI tools, which will result in hesitation to integrate AI in education. This aligns with the concerns raised by Younas et al. about how the effectiveness of AI depends heavily on sufficient internet infrastructure and institutional support. As AI becomes more integrated into education, unequal access to the internet may reinforce existing disparities in academia.
2. Updated Institutional policies are necessary for effective AI integration.
Institutional interventions should include these nuances of both positive and negative effects AI adoption will introduce, staying away from blanket bans on AI. Instead, institutions should invest in AI literacy training, update academic integrity policies to include AI usage, and teach students and educators how to use AI as a complementary learning tool instead of a replacement for teachers. As Giannakos et al. explain, AI tools present significant opportunities to enhance learning, but they also introduce ethical and pedagogical challenges that need careful institutional regulation. Similarly, a systematic review of AI applications in higher education highlights that most of the existing research is concentrated in computer science and STEM disciplines, with limited critical reflection on ethical and pedagogical implications and an insufficient representation of educators (Zawacki-Richter et al. 2019). Developing new explicit policies and AI literacy workshops will help ensure that AI is integrated responsibly into educational spaces while maintaining academic integrity. In addition, this will minimize the risk of overreliance on AI tools.
Potential Limitations
One of the most visible limitations is missing provincial coverage. The dataset reports 0 values for many provinces, and the map/table indicates that a large portion of Indonesian provinces are not represented (or have no computed average). This means we cannot confidently generalize our findings to the entire country, and we must limit claims to the provinces and regions with usable data. A second limitation is that some region-level averages are based on very few provinces (for example, Papua and NTB appear as single-province representations in the ATU table), which makes “regional” comparisons less stable and more like partial snapshots. Third, the ATU measure itself is based on a single survey statement about preferring AI over traditional learning methods; that captures an attitude, but it does not fully explain why students feel that way or whether preference reflects actual access, institutional policy, or learning outcomes. Finally, survey responses are self-reported and may be shaped by social desirability or differences in how students interpret “AI,” especially if the most common tools used in the dataset (like ChatGPT) dominate students’ understanding of what “AI learning” means.
Future Research
While this project provides insight into how Indonesian university students perceive the role of AI in education, there are several areas that could be explored in future research. One possible direction would be to include a larger and more balanced sample across Indonesian provinces, since some regions in the dataset have very few respondents. Future studies could also examine how students’ actual academic performance changes when AI tools are used in learning environments. In addition, researchers could investigate how students’ AI literacy and access to technology influence their perceptions of AI in education. Exploring these factors would help provide a more complete understanding of how AI may shape the future of higher education.