The potential of AI-powered Eye-Tracking in the Classroom:
Exploring Teachers’ Perspectives
October 2025
Author: BUSEYNE Siem
Contributor: ZERMIAN Francesca
Editor: ZAMMIT Rebecca
Through the EU-funded EYE-TEACH project, a group of international expert researchers is working on combining Eye-Tracking (ET) technologies with artificial intelligence (AI) to develop a tool that supports teachers in the classroom. The innovation of the EYE-TEACH tool comes from capturing students’ eye movement data as they read teacher-assigned texts, running it through a custom-built, AI-powered analytics system, and translating this raw data into actionable insights for teachers. Such insights could, for example, help identify students who may be struggling with reading and/or attention, or highlight when a text is too difficult or not engaging for students.
The EYE-TEACH tool is not being developed in a vacuum: to ensure the tool reflects real needs, we will gather input from different stakeholders throughout the project. Indeed, core to this research project is its constant intake of teachers’ feedback, ensuring that development is in line with what teachers will find useful and practical.
Our first feedback point was at the very beginning of the project, to help establish the research direction. On May 14, 2025, we held a virtual workshop, bringing together almost 60 teachers, researchers and stakeholders from primary, secondary, and tertiary education to explore their current experiences and expectations regarding the use of AI-supported eye-tracking systems. The discussions, split into breakout rooms for primary, secondary, and tertiary education, offered valuable opinions and highlighted what teachers are concerned about at this initial research stage.
The classroom current reality: setting the context
During the workshop, we asked teachers what superpower they would most like to have in their classroom when students are reading. A central theme that emerged across all groups was a desire to better understand what students experience while reading. As one primary teacher put it, she wished for the superpower to “be in their heads” to know what the students are looking at, where they are stuck, what micro-decisions they are making, and whether an assigned text is too difficult.
With no mindreading powers available, each teacher group indeed reported unique challenges in supporting students during reading. Primary school teachers noted that students often give up when faced with difficulties and are reluctant to ask for help. They highlighted gaps in foundational skills such as decoding, as well as limited vocabulary knowledge, which makes understanding texts harder. The diversity in student language backgrounds adds further complexity, as students with different levels of language proficiency require varied instruction. Secondary teachers highlighted issues such as superficial or fast reading, with their students often skimming without understanding. They also cited a lack of interest in reading, with some students engaging with the text only as much as is necessary to complete an assignment. Tertiary teachers observed that students sometimes misunderstand the tasks by focusing on irrelevant details, struggling to extract essential information from texts, and often show signs of demotivation or distraction, especially in online environments.
The workshop highlighted that monitoring students’ reading processes remains a challenge across all educational levels, but is felt most acutely in the primary and secondary classrooms. Teachers are typically relying on traditional methods like reading tests, asking questions, and looking at directly observable but often superficial behaviors, such as students speaking between themselves when they are not supposed to.
AI-Supported Eye-Tracking Tools: Initial Thoughts
These challenges discussed above sparked a discussion about how technology might help teachers gain deeper insights into students’ reading processes. Overall, teachers across all levels expressed strong interest in an AI-supported eye-tracking tool, particularly for diagnosing reading problems, offering personalized feedback, and identifying struggling students. In primary education, the emphasis was also on using such systems for diagnostic purposes, especially for new or special needs students. Secondary teachers focused on how such systems could reveal whether students reread texts, how they integrated textual and visual information, and whether they paid attention to keywords. A tertiary educator explained her interest in understanding why students get distracted while reading and whether their diversions were relevant or irrelevant to the task.
What should the system do?
For students, primary and secondary educators hoped the system could support students who finished tasks early by providing additional, appropriately challenging material. Tertiary teachers, while open to AI-generated adaptive recommendations, warned against over-automation, with one participant questioning whether providing automatic support would make students too passive and remove the teacher from the learning process.
From a teacher’s perspective, most participants expressed the need for the system to provide both individual- and class-level data. Across all groups, teachers agreed on the value of identifying reading strategies and comparing how different students process the same material, with class- and individual-level indicators such as engagement, attention, reading pace, patterns, and predictions of understanding. Tertiary participants also expressed interest in tools that could pinpoint the most challenging parts of a text and suggest suitable interventions, such as recommending additional reading or viewing related media.
What should it look like?
The preferred format for receiving information varied but typically included visualizations alongside textual summaries. For instance, a primary teacher suggested using emojis for immediate student feedback during activities, whereas a secondary teacher preferred a note on a teacher device with suggestions and annotated text. Tertiary teachers favored spreadsheets and graphs to allow deeper analysis. Most teachers want a combination of different functionalities of dashboards; mirroring (e.g., to monitor students), alerting (e.g., to be alerted when there is a problem), and advising (e.g., to help making decisions).
What limits should be in place?
To facilitate the use of AI-supported eye-tracking tools, all groups expressed the need for proper training and clear guidelines to ensure the system is used effectively. Ethical concerns were also raised. Primary teachers worried about the potential for student feedback to be misused or to lower self-esteem. Secondary teachers were concerned about cognitive overload and the distraction potential of digital text features.
At all levels, it was emphasised by participants that the tool should maintain teacher autonomy, and making decisions for the student should remain the teacher’s responsibility, as it requires empathy and an understanding of the student’s background. As one tertiary teacher succinctly put it, the tool should only go
“As far as the teacher tells it to.”
Some limited autonomous support was seen favourably however, such as automating tasks like identifying trigger words in the text.
What next?
Despite differences in context across countries, language, and educational level, the discussions revealed strong commonalities, underscoring again the value of and need for the EYE-TEACH project. Teachers at all levels desire deeper insight into students’ reading processes, more effective monitoring tools, and adaptable, user-friendly systems. They also share concerns about the ethical use of data, the risks associated with the automation of teaching, and the need for robust training to support implementation – safeguards that the EYE-TEACH project is integrating within the research design itself. Successful implementation will require careful attention to design flexibility, ethical considerations, and professional development to ensure that these tools enhance rather than replace the critical human element in teaching.
With the workshops kick-starting our project’s two-way communication development process with teachers, we will soon launch a more in-depth questionnaire, available in nine languages, to gather more input from European teachers. Stay tuned for more information on how to participate!
The Padlets used to collect feedback during the workshops remain open for your input:
Primary education padlet; Secondary education padlet; Tertiary education padlet.
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them.