Research Spotlight: When ChatGPT Becomes a Mathematics Learning Partner

That question sits at the heart of Ahmad Mumtaz Anwari and Joko Soebagyo’s study, “Pemanfaatan ChatGPT sebagai Asisten Belajar Matematika: Tinjauan Kualitatif terhadap Pengalaman Siswa.” The article examines vocational high school students who were allowed by their teacher to use ChatGPT while learning mathematics, capturing both the promise and the caution of AI-supported learning in an ordinary classroom setting.

For this spotlight, we invited one of the authors, Dr. Joko Soebagyo of Mathematics Education at Universitas Muhammadiyah Prof. Dr. HAMKA, to share more about students’ experiences using ChatGPT as a mathematics learning assistant.

In plain language, this article matters because it brings the debate about AI in education down to everyday classroom practice. Students are not using ChatGPT in a single uniform way. Some treat it as an answer machine, while others use it to ask for steps, explanations, and clarification when mathematics feels difficult. That nuance matters for educators and institutional leaders. A tool that can give instant answers may reduce effort if used passively, yet the same tool can invite curiosity when students are encouraged to question, compare, and verify. The study therefore points to a practical challenge for schools, not simply whether to permit AI, but how to teach students to use it responsibly. Dr. Soebagyo reflected on the broader significance of the study: “This research underscores that the primary challenge for mathematics in the AI era is not the existence of the technology itself, but rather how we educate students to use it wisely. Teachers and school leaders must shift from merely monitoring AI usage to guiding students to remain critical thinkers who master the technology they employ.”

The article offers a balanced picture of ChatGPT in mathematics learning. Three contributions stand out for readers who want to understand how AI enters ordinary classroom routines.

• ChatGPT was often used for academic tasks and quick answers. The article reports that some students submitted problems to ChatGPT and copied responses directly, which points to a risk of surface-level completion without deeper understanding.
• Some students also used ChatGPT to understand steps and concepts. Interview excerpts show that learners sometimes asked for methods, explanations, and clarification, indicating that ChatGPT can become a starting point for conceptual support when students engage actively.
• Teacher and peer support remained essential. Students reported that ChatGPT explanations were not always sufficient, and the article argues that face-to-face instruction and group discussion remain important for helping students verify and deepen understanding.

One student captured the risk of passive use clearly when he said, “Kan kalau misalnya pakai ChatGPT buat nyari jawaban doang kan gak bikin pintar. Malah bikin males kan.” The comment gives the article practical force because it shows that students themselves can recognize the difference between using AI to avoid thinking and using it to support thinking. Dr. Soebagyo added: “The most important finding from these interviews is that students need guidance, not just prohibitions. Once they understand the boundary between ‘aiding thought’ and ‘taking over the thinking process,’ they actually want teachers to design mathematics assignments that make their interaction with AI more exploratory and meaningful.”

The authors used a descriptive qualitative case study to examine students’ experiences with ChatGPT in mathematics learning. Data came from semi-structured interviews with vocational high school students who had used ChatGPT at least once in a mathematics learning session and were willing to participate actively. The data were complex because they brought together students’ reported strategies, perceptions, learning motivation, and the social context around teacher and peer support.

The interview transcripts were analyzed thematically to identify repeated patterns, main themes, and meanings in students’ experiences. The article also states that Evidano, then known as AILYZE, was used as an assistive application for systematic theme grouping and coding, while interpretation remained part of the authors’ qualitative research process.

The authors describe transcript analysis through a thematic approach, then state that AILYZE, now Evidano, assisted with “pengelompokan tema dan pengkodean data secara sistematis,” meaning systematic theme grouping and data coding.

This is a bounded assistive role. Evidano appears as support for organizing qualitative material, not as a replacement for research design, interpretation, or scholarly judgment. The findings about quick-answer seeking, conceptual explanation, motivation, and teacher support depend on the authors’ reading of the student interviews and the educational context.

For researchers experimenting with AI-supported qualitative workflows, the article offers an example of human-led analysis with AI assistance. Dr. Soebagyo said: “AILYZE is incredibly helpful in accelerating pattern recognition and the initial grouping of voluminous interview transcripts. However, AI acts merely as a data administration assistant; interpretation, sense-making, and understanding the students’ emotional and pedagogical contexts still fully require human intuition and expertise. Researchers must remember that AI cannot replace empathy and contextual sensitivity in qualitative analysis.”

For teachers, the message is neither uncritical adoption nor blanket rejection. The study suggests that guidance matters because students need help learning when to question an AI answer, when to ask for a method, and when to seek clarification from a teacher or classmate.

For researchers, the article contributes a grounded student perspective to discussions about AI in mathematics education. It reminds readers that perceptions are mixed, practices are uneven, and the learning value of AI depends on pedagogy, not only tool capability. Dr. Soebagyo said: “The most important message from this research is that no matter how advanced the technology, it can never replace the role of sound pedagogy. Whether we are teachers, researchers, or policymakers, let us focus not on the sophistication of the AI itself, but on how we guide students to remain active, critical, and empowered thinkers when interacting with that technology.”

At its core, this research asks educators to treat AI use as a learning practice that can be shaped. When schools pair technology with thoughtful instruction, students may be better positioned to use ChatGPT as a partner for inquiry rather than a shortcut around understanding.