No 24, October 2026

As technologies and societies evolve rapidly, it is necessary to examine the representations, uses and non-uses, practices, and ethical issues associated with the integration of digital artefacts into education and training more closely. Accordingly, the 23 articles brought together in this special issue of Médiations et médiatisations offer different ways of conceptualizing what constitutes a critical approach and present pedagogical experiments and reflections on the place and role of digital technologies in school, university, and training contexts. Specifically, they highlight the issues and challenges raised by technopedagogical innovations and examine stakeholders’ representations and practices, as well as the potential or demonstrated effects of different technologies and systems on teaching and learning experiences.

The following text was written for a presentation at the ROC25 conference. It aims to contribute to the analysis of what is currently happening with AI in education, linking it to what previous technological waves have taught us. It discusses the risks (particularly with regard to the loss of authority of the texts produced) and the potential (particularly as tools for problem solving). It highlights the importance of critical approaches and the difficulty of implementing them in schools for phenomena without recognized expertise. Finally, it emphasizes the importance of developing teacher agency in terms of monitoring and innovation, which raises the issue of necessary training. These points determine what can then be taught in schools.

The gradual decline in production costs has increased the accessibility of virtual reality (VR), thereby strengthening interest in this technology across various sectors, particularly in education and training. Owing to its immersive and interactive nature, VR emerges as a promising lever for skill development in academic contexts. This article reports on a study carried out in 2025 at a Quebec university, aimed at experimentally integrating VR into different teaching contexts on a small scale, with the objective of enhancing accessibility, student persistence, and academic success. The study adopts a mixed design comparing a control group, which completed the course remotely via the Zoom platform, with an experimental group that received the same instruction within an immersive VR environment. Quantitative findings indicate a statistically significant difference in final course grades in favor of the experimental group. Qualitative data derived from focus groups corroborate these results, highlighting increased course satisfaction and a reduced intention to withdraw among participants exposed to VR.

French high school students make extensive use of generative artificial intelligence (GAI) in their schoolwork. All of them? No. This study focuses on non-users and the reasons for their non-use. It is based on the responses of 153 non-user high school students, collected during a survey conducted in France in December 2024. Using a mixed-method approach, we show, on the one hand, that variables related to ethics significantly influence the intention to use GAI and, on the other hand, that the profiles of non-users are characterized by different attitudes toward ethical issues. Finally, the qualitative analysis of the reasons for non-use draws on the concepts of necessities and facts from the problematization theoretical framework. This research contributes to work on non-use in information and communication sciences and provides original insights based on the voices of actors who have received little attention in the literature: French high school students who are non-users of AI.

This article analyzes the representations, uses, and non-uses of generative artificial intelligence (GAI) among secondary school students in France. The study is based on data from a questionnaire administered at the end of 2024 to 1,159 middle and high school students in the Grand Est region, nearly two-thirds of whom reported using GAI. As part of a critical approach to digital technology in education, the study uses a social acceptability model inspired by the work of Venkatesh, enriched by variables related to the ethics and transparency of these tools. The results show that the reported uses mainly involve a pragmatic appropriation of AI tools, geared towards homework assistance, content reformulation, and information search, most often outside of formal educational settings. These practices respond to the logic of academic efficiency and time saving, while raising issues related to cognitive autonomy and understanding the limitations of technology. Non-uses appear to be varied, linked to ethical concerns, the value placed on personal effort, or institutional fears. The article highlights tensions between informal uses and uncertain educational frameworks and emphasizes the need for critical and reflective education on GAI.

This article examines the short-term effects of a six-hour continuing professional development programme for special education teachers on their perceptions of digital technologies. The programme, organized into two three-hour sessions, combines non-digital and digital learning activities. Participants first explored unfamiliar physical tools and formulated hypotheses about their possible uses based on their properties. They then discovered an unfamiliar video-editing application through a trial-and-error activity involving the creation of a "diverted advertisement", requiring them to use the digital tool beyond its intended purpose. The study is based on a pre-test/post-test design using a self-report questionnaire with a five-point Likert scale, completed by 25 participants and analyzed using paired t-tests. The results reveal significant changes in participants' perceptions regarding understanding and use of digital technologies, confidence, learning from errors, perceived ease of use, perceived safety and benefits, and digital pedagogical expertise. No significant effects were found for ease of use in front of students and colleagues, stress management, enjoyment and openness to innovation, or perceived time savings. The discussion draws on instrumental genesis to interpret the observed changes as being compatible with a process of instrumental appropriation, while highlighting the limitations of a short-term evaluation.

In the context of accelerated digital transformation of higher education (Peraya et Peltier, 2020), educational videos have become the norm, often reproducing transmissive practices that keep learners in a passive role (Fyfield et al., 2022). This contribution examines the conditions that enable teacher trainers to move beyond this technology-centred approach towards a thoughtful integration of technology and pedagogy within specific disciplines. Based on doctoral research (Ramillon, 2024) carried out in French-speaking Switzerland with seven trainers (HEP Valais, HEP Fribourg, University of Geneva), the study articulates the TPaCK model (Koehler et Mishra, 2008) and the principles of multimedia learning (Mayer et Fiorella, 2022). The methodology triangulates pre/post questionnaires, analysis of multimedia video design evolution, and semi-structured interviews. The results reveal significant progress in the TPK and TPCK dimensions, accompanied by a notable improvement in multimedia design. Beyond these developments, the study highlights critical and ethical issues: the gap between institutional requirements and the actual design conditions, as well as the need for time to develop effective educational scenarios. The article discusses how training in educational design can restore trainer agency in the face of techno-enthusiastic discourse and favour reasoned rather than hasty integration.

This article examines digital sobriety (DS) practices as responses to the environmental challenges posed by digital technology (Flipo et al., 2013) in higher education and research (HER) in France. Indeed, HER is experiencing tension: while "calls for digitisation" (Boboc and Metzger, 2023) issued by public policies promote digital technology as a driver of educational innovation, its negative environmental impacts are well known (Marquet et al., 2019). While DS appears to be a response to these impacts, its integration into practices remains unclear and its ethical dimension is rarely questioned based on the stated practices of those involved. The data presented and analysed are from 42 semi-structured interviews conducted with educational engineers (IP) and teacher-researchers (EC) between October 2024 and March 2025, as part of ongoing doctoral work in information and communication sciences. From a situated ethics perspective, we analyse the ethical dimension of DS for ECs and IPs in a professional context, as well as the potential reconfigurations of digital practices that it induces. The analysis shows that there is no single ethical framework for digital sobriety, but rather a plurality of situated digital sobrieties linked to the specific challenges faced by each individual in their professional context.

This paper investigates the acceptability of AI among a sample of nine upper secondary school teachers (teaching students aged 15 to 19) from the cantons of Vaud and Fribourg in French-speaking Switzerland. It first outlines the challenges posed by AI for the transmission of literary knowledge in school settings, with a particular focus on literature teaching. Drawing on a model of teachers’ appropriation of digital tools, the study then analyzes data from a preliminary questionnaire administered prior to a collaborative research project. The findings highlight a diversity of teachers’ conceptions regarding the integration of AI into analytical reading practices. In relation to the didactic aims of analytical reading, participating teachers define the conditions for legitimate and acceptable AI use in tasks that maintain the teacher’s mediating role, support the development of students’ critical thinking, and foster peer interaction as an essential component of literary interpretation in school-based literature education.

The rise of generative artificial intelligence systems is profoundly transforming pedagogical practices in higher education, where these tools increasingly act as mediators of knowledge. This article examines how AI use in university contexts reconfigures pedagogical mediation and the teacher–learner relationship by assuming functions traditionally associated with teaching, such as rephrasing, clarification, and learning guidance. Adopting a mediological and critical perspective on digital education, the study analyzes the effects of algorithmic mediation on role distribution, knowledge legitimacy, and student autonomy. It is based on a qualitative interpretive methodology conducted at Mohammed Premier University, combining semi-structured interviews with higher education instructors and a questionnaire administered to students from various academic levels and disciplines. The findings reveal AI-assisted student autonomy correlates with instrumental dependence, alongside a redefinition of the teacher’s role as an epistemological and critical mediator, reshaping the traditional pedagogical process. The study concludes that AI does not replace pedagogical mediation but reconfigures it by introducing a technical third party requiring clearly defined pedagogical, ethical, and institutional frameworks.

Generative artificial intelligence (GAI) can support the evaluative judgment exercised by university faculty when assessing students’ written productions. However, its use raises issues that go beyond score concordance alone. To better understand the conditions under which GAI can provide such support, this article presents an exploratory experiment carried out in a university course focused on the mastery of written French. The study compares human evaluation and GAI‑supported evaluation of the same diagnostically oriented written task. The analysis is grounded in a three‑axis framework of evaluative judgment, encompassing the functions of assessment, its stages, and its ethical dimension, understood as inseparable from the evaluator’s responsibility. The results show an overall convergence between the two modes of evaluation in terms of scores and number of identified errors, while revealing significant discrepancies in error identification, categorization, and the sequencing of diagnosis, scoring, and feedback as error complexity increases. The study shows that GAI cannot replace evaluative judgment. Nevertheless, it can support certain evaluative operations under clear methodological, institutional, and ethical guidelines, combined with faculty training and ongoing professional vigilance. It also opens avenues for further research, notably through comparison between different generative models.

The French national education system produces and collects a large volume of administrative and academic digital data concerning pupils, families and teachers. A system for governing this data, which is organized and processed by various departments, stakeholders and platforms, is being put in place. An analysis of official discourse, supplemented by a survey of interviews on data-driven management, highlights the different stakeholders’ perceptions of this data. These perceptions draw on reference systems that are not uniform. Yet, they are situated within a discursive framework that appears consistent in the central role played by the strategic use of data in harnessing digital technology for educational policies. The analysis reveals the tensions and contradictions in the implementation of data-driven management of the education system.

This article analyzes the Education Dashboard developed by the Québec Ministry of Education (MEQ) in a context where digital data are playing an increasingly prominent role in school governance. It draws on insights from the sociology of quantification, critical data studies, and the theory of public policy instruments. The analysis is based on a qualitative, documentary, and interface-centered approach, complemented by an examination of the strategic documents framing the dashboard. The findings show that the dashboard privileges a representation of the education system centered on narrowly measurable dimensions, such as graduation, performance on ministerial examinations, enrolment, and resources, while relegating pedagogical processes, actual educational work, and local contexts to the background. They also highlight the structuring role of comparability, the risks of implicit rankings, and a tendency toward the symbolic recentralization of educational steering through data. The findings further indicate that, despite the heterogeneity of reference dates and data consolidation statuses, the dashboard interface tends to present these data as synchronous and comparable, thereby contributing to the naturalization of their timeliness and reliability, both of which remain open to question.

The learning experience refers to the perceptions and outcomes resulting from interactions between learners and their environment. In the context of higher education, the integration of serious games introduces more interactive, engaging, and experience-centered learning modalities. The augmented reality serious game Pangu, designed to support the learning of biological concepts at the microscopic level, has already been evaluated for its design quality, but without empirically analyzing the students' learning experience. The objective of this study is to evaluate the learning experience associated with this tool and to examine the extent to which it reaches the intended learning objectives. An experiment was carried out with 18 students who completed a knowledge questionnaire before and after using the game as well as a questionnaire evaluating the learning experience at the end of the experiment. The results show a significant improvement in student performance and highlight the positive impact of factors such as playfulness, objective clarity, usability, and cognitive development associated with learner engagement. These findings invite further reflection on the role of digital game-based tools in a university context as levers for active learning, while also raising questions concerning the conditions necessary for their integration into teaching.

This article offers an information and communication analysis of sexuality education in the digital age (e-SexEd) based on a survey of adolescents, parents, and professionals involved in such programmes. It examines the conditions for the legitimization, implementation and appropriation of sexuality education in media environments that have been profoundly reconfigured by digital technology. The analysis highlights three structural tensions: institutional legitimacy weakened by often alarmist or legal discourses; school programme fragmentation marked by ad hoc interventions that are not well integrated into the educational timetable; and the central role of digital technology in adolescents’ practices for accessing information on sexuality. Far from reflecting a rejection of sexuality education, the digital practices observed testify to a strong demand for resources, narratives and discussion spaces that enable the appropriation of sexual knowledge in a specific context. Based on these results, the article proposes an infocommunicational definition of e-SexEd, conceived as a system that structures the production, circulation and reception of educational discourse within hybrid media environments, shaped by multiple regimes of legitimacy and differentiated appropriation logics among adolescents.

Recent studies show that the introduction of AI into teaching can generate psychological tensions or professional concerns, particularly when it is perceived as a threat to teachers’ role or expertise, rather than a complementary resource. Drawing on activity theory and affect theory, this exploratory study aimed to understand the reconfiguration of teaching activities from the perspective of future teachers. A questionnaire was administered to 86 pre-service teachers, covering ten typical activities. The results reveal four activity configurations, independent of the participants’ comfort with or attachment to these activities, but more related to the place of the activity in the profession.

This study evaluates a hybrid teaching method used in the third year of a bachelor’s degree in education sciences, combining a lecture on “digital learning environments” and ten tutorials focused on the collective creation of an e-portfolio, including an individual reflective component. Designed to develop collaborative skills, the program aims to empower students through techno-pedagogical hybridization. It is based on a shared Moodle space offering accessible and inclusive resources and explores its innovative and emancipatory potential through the combinations of lecture- tutorials, presence-distance, and evaluation-co-evaluation (Maturana et Varela, 2012). The study, carried out with two teaching assistants and 276 students, combines a teacher focus group, a student questionnaire, and e-portfolio analysis (n=69). The results show that attendance is crucial in both lectures and tutorials: students who attend develop research skills (problematization, analysis) and psychosocial skills (problem solving, communication), moving from a state of wonder to a critical and autonomous stance. Conversely, those who are absent, often struggling to link lectures and tutorials, resort to circumvention strategies, hindering any empowerment process.

This quasi-experimental study examines the effectiveness of two serious investigative digital games (Genially and Unity), compared with a no-game control group, in supporting college-level English as a Second Language (ESL) learning of concepts related to digital literacy and artificial intelligence. Learning is operationalized through two components of the form–meaning association of a word: 1) written form recall (productive, written) and 2) meaning recognition (receptive, written), measured with a pre-test and a post-test. Questionnaires were administered using LimeSurvey. Statistical analyses (N = 189; A = 53, B = 71, C = 65) were conducted using ANCOVA while controlling for baseline performance, with Holm-adjusted post-hoc comparisons (Holm method). Results show no significant differences between groups for meaning recognition. By contrast, a group effect is observed for recall of the written form, a noteworthy finding since this component is more difficult to master than meaning recognition. The Genially group thus achieved higher adjusted post-test scores than the control group and the Unity group. These findings suggest that the impact of serious games on learning may vary depending on the component assessed, learners’ prior knowledge, and intervention design.

With the rapid development of digital education in recent years, teachers today are expected to incorporate them as a theme into their classroom activities. From introductions to programming to media literacy and discovering digital uses, some teachers are therefore in the process of completely transforming their practices. At the same time, they are also being asked to educate their pupils about sustainability, even though the negative impact of digital technology in this area, and particularly artificial intelligence, is increasingly being highlighted. So, how are teachers coping with these two seemingly contradictory demands? And how do they manage to position themselves and contemplate their personal and professional future in a world in which digital technology has taken on such an important role? In other words, are we currently training teachers who are resistant to digital technology or who are techno-solutionists? This connection between digital technology and sustainability will therefore be the focus of this article. To answer this question, we carried out a study on the digital practices of pre-service teachers and their vision of the future, which we present in this article, with the aim of understanding their worldview in order to better guide them in supporting their students.

Over the past three years, the digital landscape has been profoundly transformed by the rapid and widespread emergence of generative artificial intelligence (GAI) technologies. Their integration into the field of education—for both students and teachers—now appears inevitable. Building on previous research (Bruillard, 2025), this article, drawing on a study carried out with postsecondary students, offers an analysis of the process through which individuals develop their capacity to act, learn, and empower themselves within their environment (empowerment dynamic) to explain the uses and non-uses of GAI technologies. From an intersectional perspective, various factors of conversion and choice are identified to understand the emergence of a realm of possibilities and its realization by students based on a set of factors. These analyses lead two findings. AI technologies are both a problem and a solution, calling for a shift away from an oppositional approach toward a comprehensive understanding of enabling dynamics and potential for social innovation. This requires addressing the issue of AI’s relationship with education through the lens of multiple tensions, particularly temporal, cognitive, metacognitive, and evaluative. The importance of a clear usable framework is emphasized as a means of reducing potential sources of stress and supporting the development of academic skills aligned with the demands of the socio-economic world.

This reflective account of practice explores the implementation of conditional progression learning (CPL) in high school mathematics education within the Quebec context. CPL is based on the idea that students must master each step before progressing, which poses significant challenges in heterogeneous classrooms. The author describes the limitations of traditional approaches and the difficulties encountered with common digital tools (Google Forms, Microsoft Teams), which do not allow for effective feedback or automated conditional progression. After several experiments, an innovative solution was developed: the use of PDF forms programmed in JavaScript, integrated into Teams. This method ensures real-time answer validation, progress saving, and automated grading, reducing the time allocated to correcting simple exercises. This allows the teacher to focus interventions on mathematical reasoning and fosters greater student autonomy. Although this approach requires considerable initial time and technical skills investment, it represents a significant advancement for the individualization of learning. The author concludes by highlighting the potential of technologies, notably artificial intelligence, to create integrated and accessible tools that would support CPL and contribute to reducing teacher burnout.

This article critically analyzes the integration of immersive virtual reality into a collaborative learning system in a university setting. It is based on a pedagogical experiment carried out with 42 undergraduate students as part of a course on mediated collaboration and addresses three challenges identified in the literature: the pedagogical gap associated with the lack of established practices and recommendations, the constraints of social interaction in an immersive environment, and the technical and cognitive barriers that can affect learning. Workshops, structured in three immersive cycles, mobilized differentiated roles around a three-dimensional co-construction task using the Arkio software. The study adopts an exploratory qualitative approach based on in situ observations and field notes produced by the research team, analyzed using an inductive approach structured around three analytical axes and supported by data triangulation. The results show that the success of the workshops depends mainly on the quality of the educational scripting and logistical orchestration, including technical preparation, instruction structuring, shared visual feedback improvement, and gradual role adjustment during the activity.

Critical approaches to digital technology in education and training seem to be gaining popularity. Perhaps international geopolitics and the state of the world are making it easier to move beyond technocentric approaches focused on added value and instead embrace approaches that are sensitive to living things. But what constitutes a critique of digital technology in education? How can work in this field be characterized? Drawing on sociology and recent work in education and training, this article aims to propose a naive framework—one that therefore deserves to be critiqued and tested—to ensure that the critique is well formulated. Indeed, the disciplinary field of digital technology in education is broad, and it is neither possible nor desirable to reduce it to the level of methodology or objects of study. So, what brings together critical approaches to digital technology? Values as the entry point is consistent and seems to allow most of the work encountered to be situated.

Our digital world is once again being upended by the emergence and rapid development of artificial intelligence available to the general public. This is affecting software development, including software dedicated to education. But it is not yet clear what this implies in terms of training people or in terms of ethics. This paper examines the specific case of developing software dedicated to prior learning recognition. What do we already know about artificial intelligence? What can we deduce about its possible consequences? This reflection highlights many issues ranging from ecology to expertise loss o in general and focuses more specifically on the foundations of prior learning recognition and how these could be affected by artificial intelligence.