The focus of this special issue is on the design, implementation, and evaluation of trustworthy AI-enhanced educational systems, with particular emphasis on the three interconnected pillars of data governance, pedagogical evidence, and human oversight. As AI-mediated tools—from personalized learning recommendations to automated formative feedback—are increasingly introduced across educational settings and cultural contexts, they may improve instructional efficiency, personalization, and responsiveness (Cai et al., 2026; Lee et al., 2023). At the same time, their adoption raises concerns about opaque algorithmic decision-making, insufficiently transparent educational-data practices, uneven empirical validation of pedagogical claims, and risks to educator autonomy when algorithmic outputs are treated as authoritative (Alfredo et al., 2024; Gillani et al., 2023; Zhu et al., 2025). These are not solely technical problems but socio-technical, pedagogical, and ethical challenges that require coordinated attention to policy, institutional practice, system design, educational implementation, and evaluation (Holmes et al., 2022; Nguyen et al., 2023). The special issue therefore examines how trustworthy AI-enhanced educational systems can support learners and educators while preserving learner and educator agency, accountability, and professional judgment (Molenaar, 2022).

The scope of this special issue encompasses empirical, design-based, theoretical, and review studies that address these challenges through an integrated educational and socio-technical lens. It recognizes learners and educators as active stakeholders, with educators retaining responsibility for pedagogical decisions and learners being afforded appropriate transparency, agency, and opportunities to question or contest consequential AI-supported decisions (Alfredo et al., 2024; Holstein et al., 2019). We invite contributions that develop and evaluate education-specific data governance approaches, including data minimization, privacy-enhancing technologies, access controls, data provenance and documentation, consent, and accountability mechanisms (Liu & Khalil, 2023; Nguyen et al., 2023). We also welcome studies that establish pedagogically grounded evidence of AI effectiveness through rigorous experimental, quasi-experimental, longitudinal, design-based, learning-analytics, or mixed-methods research (Chen & Cheung, 2025; Deng et al., 2025; Zhao et al., 2023), as well as studies examining whether and under what conditions explainable AI supports understanding, calibrated trust, and informed educational decision-making (Khosravi et al., 2022). A further area of interest is the design and evaluation of meaningful human oversight across the AI system lifecycle, including mechanisms for review, intervention, contestation, escalation, and accountability where AI-generated recommendations affect teaching, learning, or assessment (Molenaar, 2022; Mouta et al., 2025). Additional topics include professional learning and AI literacy for informed oversight, institutional governance and ethics auditing, algorithmic fairness and bias assessment, and the appropriate balance between automated support and professional pedagogical judgment (Idowu, 2024; Wang et al., 2024). A cross-cutting theme is the integration of data governance, pedagogical evidence, and human oversight into coherent frameworks, design principles, and scalable practices evaluated in authentic educational settings (Majumdar et al., 2024; Niu et al., 2026).

The rationale for this special issue lies in the need to connect research areas that are often examined separately but are closely interdependent in educational practice (Alfredo et al., 2024; Li & Gu, 2023). Responsible data governance supports the legitimate, secure, and accountable use of educational data (Liu & Khalil, 2023); rigorous pedagogical evidence is needed to determine whether AI-supported interventions improve teaching, learning, or assessment (Boolzen et al., 2026; Ma & Zhong, 2025); and meaningful human oversight enables educators, learners, and institutions to interpret, question, and act appropriately on AI-generated outputs (Holstein et al., 2019; Molenaar, 2022). These three pillars are closely interdependent: responsible data governance helps establish the conditions under which pedagogical evidence can be generated, interpreted, and trusted; pedagogical evidence, in turn, informs what forms of human oversight are most needed and where; and human oversight helps ensure that both data practices and evidential claims remain accountable to educational values and contextual realities. Without coordination among these elements, educational AI systems may be adopted without adequate privacy safeguards, demonstrable educational value, or clear lines of responsibility and accountability (Yan et al., 2024; Zhu et al., 2025). By providing a dedicated interdisciplinary forum, this special issue aims to advance empirically grounded, ethically responsible, and pedagogically meaningful applications of AI in education. It will also seek to generate actionable design principles, evaluation methods, and governance frameworks that can inform researchers, practitioners, institutional leaders, system developers, and policymakers working toward trustworthy AI-enhanced education.

The proposed special issue is differentiated from general collections on AI in education by treating data governance, pedagogical evidence, and human oversight as interdependent conditions for trustworthy educational AI rather than as separate ethical, technical, or pedagogical topics (Alfredo et al., 2024; Holmes et al., 2022; Li & Gu, 2023).

Individual submissions will not be required to address all three pillars equally. However, each manuscript should examine at least one pillar in depth and explain its implications for at least one of the other pillars. For example, a study of an explainable assessment system should consider not only predictive performance but also the quality of the pedagogical evidence, the data used to generate the assessment, and the ability of educators or learners to review and question consequential outputs (Niu et al., 2026; Oliveira et al., 2026).

To facilitate comparison and synthesis across the accepted papers, authors will be encouraged to report the following elements where applicable:

• The educational context, target users, and intended pedagogical purpose;

• The function of the AI system and the consequences of its outputs;

• The origin, provenance, quality, and governance of the data;

• The pedagogical theory, intervention logic, or learning mechanism;

• The educational processes and outcomes evaluated;

• The allocation of responsibilities among educators, learners, institutions, developers, and AI systems;

• The mechanisms available for explanation, review, intervention, correction, or contestation;

• Relevant privacy, fairness, accessibility, security, and accountability considerations; and

• Limitations concerning generalizability, transferability, and implementation at scale.

Priority will be given to research conducted in authentic educational settings and to studies that evaluate educational processes or outcomes over a sufficient period. Rigorous design-based research, experiments, quasi-experiments, longitudinal studies, mixed-methods investigations, comparative case studies, and systematic evidence syntheses will be welcomed. Conceptual and theoretical contributions may also be considered where they provide a clearly defined and analytically rigorous framework that advances the educational study of trustworthy AI (Deng et al., 2025).

Responsible research transparency will be encouraged without requiring the public release of sensitive learner data. Where legal, ethical, or institutional restrictions prevent data sharing, authors may instead provide appropriately anonymized metadata, analysis code, research instruments, prompts, system documentation, synthetic examples, study protocols, or detailed methodological descriptions (Liu & Khalil, 2023; Yan et al., 2024).

The guest editors will synthesize the accepted papers using the three-pillar framework and examine how governance arrangements, evidential standards, and forms of human oversight interact across the design, deployment, evaluation, and continuing use of educational AI systems. It will identify areas of convergence, unresolved methodological issues, and priorities for future research rather than presenting the papers as an unrelated collection of AI applications.

Through these measures, the special issue is intended to provide future researchers with a structured body of evidence, comparable reporting dimensions, and clearly identified research gaps concerning the conditions under which AI can be responsibly and educationally integrated into teaching, learning, and assessment.

Click here for more information

Guest Editors:

Niels Pinkwart (corresponding guest editor)

Humboldt University of Berlin

pinkwart@hu-berlin.de

Olga Viberg

KTH Royal Institute of Technology

oviberg@kth.se

Jingsha He

Beijing University of Technology

jsh.bjut@gmail.com

Haoran Xie

Lingnan University

hrxie@ln.edu.hk

Chee-Kit Looi

The Education University of Hong Kong

cklooi@eduhk.hk

The proposed special issue, “Sustainable Utilization of Educational Data for Learning and Teaching,” focuses on how educational data can be used sustainably and responsibly to facilitate educational practice. It encourages data practices that preserve educational value over time, remain meaningful across contexts, and support informed use by different stakeholders. Here, sustainable utilization refers to the long-term availability of data and their continued value for intelligent support of teaching and learning, in line with the core direction of SDG 4 (“ensure inclusive and equitable quality education and promote lifelong learning opportunities for all”).

The issue is timely as educational institutions increasingly rely on diverse forms of data to inform instruction, learner support, assessment, and educational decision-making. The permeation of AI (and generative AI) into society has made data-driven educational technologies more important to examine not only what can be done with educational data, but how such data use can be made pedagogically meaningful. This direction follows the rapid growth of Large Language Models (LLMs) related educational research (Shi et al., 2025), the increasing attention to explainable AI in educational settings (Khosravi et al., 2022), and recent efforts to co-design data-driven educational technology with practitioners (Ogata et al., 2024).

Educational data now extends beyond conventional learning logs. Current intelligent learning systems increasingly draw on multimodal sources, including textual, behavioral, audiovisual, sensor-based, physiological, and interaction data from assorted learning environments. As Oviatt (2022) notes, multimodal interaction and analytics increasingly incorporate a greater number of information sources, creating both richer opportunities for educational data use and more demanding requirements for data integration, interpretation, and system design. This makes sustainable utilization not only of data access but also of diverse modalities, processed in ways that support reproducibility, transferability, and continued analytical value across diverse educational settings.

Apart from system design and technical infrastructures, a responsible approach to educational data utilization also requires attention to privacy, governance, transparency, fairness, and inclusion. These issues become especially important when educational data is reused over time or across contexts, and when data-informed systems affect learners from diverse cultural and institutional backgrounds. For instance, Ito et al. (2025) discuss risk-based privacy protection, highlighting the tension between preserving analytical value and reducing re-identification risk in data sharing of sensitive educational data. Focusing on recent policy initiatives in Switzerland and Japan, Hangartner et al. (2025) show that data sharing for student-centered education is increasingly tied to new forms of data-based governance that connect classrooms, research, and system-level decision-making. This makes governance, access, and responsible data use central issues in any discussion of sustainable educational data utilization.

By bringing together work from sustainable and responsible educational data use, this special issue aims to advance evidence-based work on how educational data can support teaching and learning in ways that remain meaningful, trustworthy, and reusable over time.

We welcome submissions including, but not limited to, the following topics:

• Empirical studies utilizing sustainable educational data from authentic educational settings to examine how such data inform instructional design, learner support, feedback, or assessment practices.

• AI, generative AI, and learning analytics that facilitate the utilization and interpretation of sustainable educational data in educational practice.

• Studies on the use, integration, or interpretation of multimodal educational data in diverse educational settings.

• Cross-context reuse, longitudinal use, or secondary use of educational data for pedagogically meaningful purposes.

• Ethical, privacy-aware, fair, and inclusive approaches to the collection, management, sharing, and use of sustainable educational data in authentic educational settings.

• Educational technologies, data workflows, or analytic processes that are grounded in sustainable educational data and improve learning and teaching.

Click here for more information

Guest Editors:

Guandong Xu

Education University of Hong Kong

gdxu@eduhk.hk

Oscar Lin

Athabasca University

oscarl@athabascau.ca

Jing Lei

Syracuse University

jlei@syr.edu

Changhao Liang

Kyushu University

bluster3a@gmail.com

Chengjiu Yin (corresponding guest editor)

Kyushu University

yin.chengjiu.247@m.kyushu-u.ac.jp

We invite empirical, methodological and design-oriented studies that identify, explain, and respond to heterogeneous learner profiles in GenAI-supported education, moving beyond designs that treat learners as a single homogeneous group. Relevant settings include STEM, academic writing, literacy, humanities, social sciences, health professions, business, teacher education, vocational education, arts and design, and interdisciplinary learning. We particularly welcome person-centered and process-oriented methods such as latent profile and cluster analysis, growth mixture and latent transition modeling, sequence analysis, process mining, epistemic network analysis, multilevel person-centered modeling, learning analytics, qualitative comparative analysis, case-based designs, and mixed methods. Suitable studies might examine heterogeneous patterns of GenAI use, prompting, feedback uptake, help-seeking, verification, revision, collaboration, self-regulation, metacognitive monitoring, AI literacy, and learning outcomes over time. Work that develops learner typologies, maps engagement pathways, models differential effects, or designs adaptive scaffolds for diverse learners is especially encouraged.

Specifically, this special issue welcomes contributions that address, but are not limited to, the following topics:

• Learner profiles in GenAI-supported education and the differential effects of GenAI across learner groups

• Engagement pathways and longitudinal trajectories of GenAI-supported learning

• GenAI and educational equity, including the reproduction or mitigation of achievement gaps

• GenAI in STEM, humanities and social sciences, and professional and vocational education

• Prompting practices, feedback uptake, and adaptive learning in relation to learner heterogeneity

• Self-regulation and metacognition in GenAI use

• Overreliance, cognitive outsourcing, and productive struggle in GenAI use

• Adaptive scaffolding and pedagogical design for diverse learner groups

• Methodological advances in person-centered GenAI research

• Teacher, peer, and institutional mediation of GenAI use

• Ethical and responsible GenAI use among diverse learners

Click here for more information

Guest Editors:

Michael Yi-Chao JIANG

Shenzhen Technology University

mjiang@sztu.edu.cn

Morris Siu-Yung JONG (corresponding guest editor)

The Chinese University of Hong Kong

mjong@cuhk.edu.hk

Xiaoming ZHAI

University of Georgia

Xiaoming.Zhai@uga.edu

Please Note:

• Educational Technology & Society is a fully open-access journal and does not charge publication fees or article processing charges (APCs).

• Any message requesting payment in exchange for acceptance or publication is fraudulent.

• The ET&S editorial office will never request payment from authors, reviewers, or readers.

• Official communications are sent only through the ET&S editorial system or verified addresses listed on our website.

What You Should Do:

• Do not respond to suspicious emails and never send money.

• Report fraudulent messages immediately to the ET&S Editorial Office at ets.submit@gmail.com.

• Always confirm editorial correspondence through the official ET&S website: http://www.j-ets.net.

Educational Technology & Society (ET&S) welcomes special issue proposals on specific themes or topics that address the usage of technology for pedagogical purposes, particularly those reflecting current research trends through in-depth research. 

For more information, please visit the Special Issue Proposals page.

The ET&S Editorial Office