Artificial Intelligence in Inclusive Education

Edited by Jan-René Schluchter¹ and Jakob Sponholz²

¹ Ludwigsburg University of Education, Germany

² University of Cologne, Germany

Please submit your abstract until 19 January 2026 at https://www.medienpaed.com/about/submissions. Please also find the author guidelines there.

Call for Papers as PDF

In recent years, Artificial Intelligence (AI) has increasingly found its way into a multitude of societal domains – not least in the education sector. AI is an umbrella term for various technologies that replicate human-like problem-solving abilities. One key area is machine learning. Unlike traditional software development, where every step must be explicitly specified, machine learning systems can learn from examples and recognize patterns (Samuel 1959). In recent years, generative AI systems in particular have changed the way content is created:

“Generative AI is an umbrella term for AI applications that often draw on a whole series of AI methods, combine them, and use large amounts of data as a basis (e.g. large language models). The goal of generative AI applications is to produce all kinds of results and products, for example texts, images, and videos.” (Scheiter u. a. 2025, 7; [translated])

The implementation of artificial intelligence in educational contexts thus constitutes a multifaceted challenge that intersects with scholarly debates on digital inequality (Schluchter 2023), digital inclusion (Bosse, Schluchter, and Zorn 2019), and digital participation (Bosse and Sponholz 2023). The focus is on the critical analysis of the design, but also the possibilities for designing AI systems and their integration into societies in general and the education system in particular:

“As we navigate the complexities of this digital transformation, it’s crucial not just to celebrate advancements but also to critically examine who might be left behind. (...) This is not just about equitable access to technology but about empowering all individuals with the knowledge and skills to thrive in an AI-driven world.” (Hendawy 2024)

The European Commission has also addressed the ethical use of AI in educational contexts. In the framework paper “Ethical Guidelines for Educators on the Use of AI and Data for Teaching and Learning Purposes”, four perspectives on the use of AI in school contexts are distinguished: (1) Student Teaching (2) Student Supporting (3) Teachers Supporting, and (4) System Supporting (European Commission 2022, 14).

With regard to the potential of AI for inclusive education, it is clear that learner-centred AI systems and applications such as intelligent tutoring systems (ITS) or adaptive learning platforms can contribute to the individualisation of learning processes and thus take into account the individual needs of students and the diverse learning requirements (Khan 2024; Varsik and Vosberg 2024). At the same time, this opens up the possibility of improving the accessibility of content for students: AI systems can be used to translate texts into simple language, generate alternative texts for images and adapt learning content to individual needs (Sponholz and Wolf 2025; Varsik and Vosberg 2024).

However, these opportunities also entail risks for pupils’ participation in education, as the use of AI must be viewed in the context of existing exclusions (such as unequal access to AI systems and varying abilities and skills in dealing with AI), the individual suitability of AI applications and pupils, and the respective contexts of schools and teaching – and thus creates both potential and limitations for equal opportunities. If, for example, AI is used to support pupils in a performance situation, its use must be carefully considered:

“If AI [in this context: LLMs] is used to compensate for disadvantages in examination situations, the ethical assessment of its legitimate use becomes even more challenging. Here, the new opportunities for participation are directly opposed to the challenges. [...] However, these entirely new opportunities raise the question of whether the use of such technologies in the context of compensating for disadvantages might not be so effective that it exceeds the goal t o reach equity. If the goal of equity is indeed exceeded, then this could be described as overcompensatory use, i.e. a form of support that, in an effort to achieve greater equity for the individual, creates new inequalities within the group of learners.” (Sponholz and Wolf 2025, 350; [translated])

In addition to this example, there are many other ways in which AI can be used in the everyday professional lives of teachers to support not only students but also teachers themselves. This includes not only the differentiation of content, but also the planning of lessons, research and even teachers’ own professional development (Kasneci et al. 2023). Teachers can also benefit from AI-based training and assistance systems that support pedagogical decision-making processes and facilitate differentiation in the classroom (as cybernetic teammates) (Dell’Acqua et al. 2025).

At the institutional level, AI offers opportunities in the areas of diagnostics and individual support. This may facilitate more targeted resource allocation, thereby enabling preventive approaches to potential educational disadvantages. Another key aspect is the professionalisation of teachers: the effective and responsible use of AI tools requires specific skills that often need to be systematically developed in initial and continuing training (Kasneci et al. 2023).

In current discourse on these applications in teaching and learning contexts, it is pointed out on the one hand that both pupils and teachers can become dependent on AI models and, for example, fail to fully master their problem-solving skills:

“The effortlessly generated information could negatively impact their critical thinking and problem-solving skills. This is because the model simplifies the acquisition of answers or information, which can amplify laziness and counteract the learners’ interest to conduct their own investigations and come to their own conclusions or solutions.” (Kasneci u. a. 2023, 5)

In order to make AI truly productive for learning processes, tasks must be designed in such a way that they do not lead to cognitive offloading, i.e. the outsourcing of the cognitive capacities and skills required for a task to AI (Gerlich 2025). But there are also risks for teachers in regarding their own profession:

“Using large language models can provide accurate and relevant information, but they cannot replace the creativity, critical thinking, and problem-solving skills that are developed through human instruction. It is therefore important for teachers to use these models as a supplement to their instruction, rather than a replacement.” (Kasneci u. a. 2023, 6)

On the other hand, current discourses on AI and education in this context point to the relationship between humans and machines, or human abilities in relation to machine abilities. For example, AI (ChatGPT) completed the Torrance Test of Creative Thinking, which was designed to measure human creativity (Guzik, Byrge, and Gilde 2023). Discussions about human-AI complementarity are also proving to be relevant, not least in the context of inclusive education – both for pupils and teachers (Melo-López et al. 2025; Holstein and Aleven 2022).

However, the potential for participation through AI can only be realised if ethical, social and technological challenges are consistently addressed and all the factors governing the use of AI interact productively (Sponholz and Wolf 2025). It is therefore crucial to integrate AI responsibly, focusing on learners with their individual needs and the principles of equal opportunities and inclusion (Autenrieth, Schluchter, and Schulz 2025).

Viewed through the framework of inclusion theory, questions of AI safety (Bengio et al. 2025) and AI alignment (Leike and Sutskever 2023) emerge as central concerns: How can AI systems be designed to ensure not only technical safety but also to actively counteract mechanisms of exclusion? How can we prevent certain values and power structures from being permanently ‘frozen’ by AI systems (value lock-in) (MacAskill 2022)?

These questions are closely related to inequality-relevant structures within AI systems themselves. Training data reproduces societal biases, and algorithmic decision-making processes follow normative assumptions in society: for example, in the form of normative performance expectations that systematically exclude certain students based on their dispositions, such as students with disabilities or impairments (Varsik and Vosberg 2024).

Similarly, many AI systems currently used in schools and classrooms are the responsibility of (a small number of) commercial enterprises, which creates a risk of dependency on these actors and, at the same time, allows commercial actors with their profit interests (but also their worldviews and views of humanity) to exert increasing influence in the education sector. This leads not only to the reproduction of existing mechanisms of inequality, but also to the emergence of new inequalities – especially in the context of inclusion and inclusive education (AI Divide) (Carter, Liu, and Cantrell 2020).

This special issue aims to systematically explore the tension between the technological possibilities and social power relations of AI in the context of inclusive education, with a focus on schools and teaching, and to develop perspectives for a critically reflective design of these transformation processes – with a particular focus on the significance of media education.

We invite submissions of theoretical and empirical articles that address the following questions:

Current AI use in special education and inclusive settings

• Current uses of artificial intelligence models in inclusive and special education contexts – by teachers and/or students
• Assistive applications for inclusive learning – empirical contributions to improving accessibility (e.g. the perceptibility and comprehensibility of learning materials through AI)
• Professional development (e.g. studies on training and continuing education that enable teachers to use AI responsibly and design inclusive learning environments).

Pedagogical and didactic concepts

• How can (media) educational programmes and measures be designed to encourage critical reflection on AI and inclusion?
• Which (media) didactic approaches reflect connections between AI and inclusive education or inclusive didactics?
• How can AI be addressed in the classroom to enable pupils to actively reflect on the implementation of AI in society?

Assistive technologies and (digital) accessibility

• Which AI-supported assistive technologies exist for various forms of impairments (visual, auditory, cognitive, motor)?
• How can AI tools for translation into plain language/Easy Language be evaluated and improved?
• How can the quality of automatically generated alternative text and subtitles be evaluated?

Adaptive learning systems and digital inequality

• What AI-supported assistive technologies exist for different types of impairments (visual, auditory, cognitive, physical)?
• How can AI tools for translating into easy language be evaluated and improved?
• How can the quality of automatically generated alternative texts and subtitles be assessed?

AI-Bias

• How do AI-based adaptive learning systems affect different groups of students? Do they contribute to reducing or reinforcing subject- or skill-related performance differences?
• What forms of the AI divide manifest themselves in the school context in relation to access to and use of AI-based adaptive learning systems?
• How can disparities in access to and use of AI tools be systematically assessed and addressed?

Techno-ableism and participatory technology development

• How does techno-ableism manifest itself in current AI applications for education?
• What role do people with disabilities play in the development of assistive AI technologies?
• How can principles of crip technoscience (Shew 2022) be implemented in the design of education-related AI tools?
• To what extent do current AI tools promote or hinder the self-determination and autonomy of learners with special needs?

Socio-emotional learning and human interaction

• What impact does the increased use of AI tools have on social interaction and relationship quality in the classroom?
• How can AI-supported learning environments be designed to promote rather than hinder socio-emotional learning? In this context, how should individual support with and without AI applications be conceived?
• What role do AI chatbots and robots play in inclusive educational contexts?

Teacher professionalization

• What AI-related skills do teachers need for inclusive teaching?
• How should in-service training be designed to enable teachers to use AI critically and reflectively?
• What disparities exist in access to AI-related training for teachers?
• How can AI tools support teachers in differentiation and individualisation without contributing to techno-solutionist approaches?

AI alignment and value lock-in

• How can AI systems be designed to align with inclusive values?
• What mechanisms prevent problematic norms and values from becoming permanently ‘frozen’ in AI systems? How can media education contribute to this?
• How can the participation of marginalised groups in the design of AI values be ensured? What potential do educational contexts in particular offer here?

We kindly ask interested parties to send an abstract (max. 500 words excluding references; PDF document) for their planned contribution no later than 19 January 2026 via the MedienPädagogik journal platform: https://www.medienpaed.com/about/submissions.

The abstracts will be reviewed and feedback provided to the authors on an ongoing (rolling) basis, so that submitters will be informed soon after their submission about the status of their suggested article. After a positive evaluation, the invited authors will be asked to submit their full texts no later than 31 May 2026.

Long articles (about 40.000 characters including spaces, excluding abstract and references) may be submitted. Contributions may be submitted in German or English. Please note the instructions for manuscript submission. Submitted texts must be original contributions or first publications.

An abstract of 150–200 words summarises the central statements and results. Both the title and abstract of the contribution must be available in German and English. Publications are open access under the CC BY 4.0 licence.

Contributions are reviewed by double-blind peer review.

Authors are kindly reminded that MedienPädagogik follows the highest ethical standards in academic publishing. In this regard, we encourage the responsible and transparent use of AI tools, as outlined in the journal’s official statement:

“MedienPädagogik is committed to maintaining the highest ethical standards in academic publishing. In line with the recommendations of the Committee on Publication Ethics (COPE), we expect authors, reviewers, and editors to use artificial intelligence (AI) tools responsibly and transparently.

• Authors shall clearly disclose the use of any generative AI tools in the writing, editing, or translation of their manuscripts. This includes AI assistance in data analysis, media generation, or literature reviews.
• AI tools cannot be credited as authors. Responsibility for the content remains solely with the human authors.
• Data protection, fairness, and the potential for bias must be considered when using AI throughout the research and publication process.
• Reviewers and editors may use AI tools to support but not replace human judgment. Editorial decisions will never be made solely by algorithms.”

https://www.medienpaed.com/about/submissions

• Submission of abstracts (max. 500 words): from now until January 19, 2026
• Feedback on abstracts: shortly after submission (rolling), by February 06, 2026
• Submission deadline for full texts: May 31, 2026
• Planned completion of peer-review: by the end of July 2026
• Planned publication: Summer/Autumn 2026