Most conversations about AI and inclusion in education start and end with disability. That’s accurate, but incomplete. A new systematic review by Navas-Bonilla et al. (2025) pulls together 159 studies on how technology supports inclusive education, and the most useful thing it does is widen the frame. Inclusion, the authors argue, covers every barrier that blocks equitable access to learning: poverty, language, culture, geography, chronic illness, age.
I’ve written before about how AI can support students with disabilities, drawing on research like Linsenmayer’s OECD report on AI and special education (2025). This review adds another layer. It maps the full technological ecosystem, from Braille displays and adapted keyboards to augmented reality and educational robotics, across every form of inclusion the literature has addressed in the last five years.
What Counts as Inclusive Education Technology Research
Navas-Bonilla et al. searched the Scopus database using keyword blocks around technology, education, and inclusion, following PRISMA guidelines. The 159 studies they analyzed span publications from 2019 to 2024.
The breakdown is revealing. Sensory disabilities dominated at 60% of the studies. Gender equality came next at 36%, followed by racial and ethnic inclusion (35%), physical disabilities (34%), and mental or intellectual disabilities (34%). Cultural inclusion appeared in 30% of studies, socioeconomic inclusion in 28%, and learning difficulties in 27%. Linguistic inclusion, age, and health conditions brought up the rear at 8%, 4%, and 2% respectively.
That distribution tells a story. The research community gravitates toward the most visible forms of exclusion, the ones with clearest technological fixes. A screen reader for a blind student is a straightforward intervention. Technology that addresses the intersection of poverty, language, and cultural difference is messier, and far fewer researchers are tackling it. Navas-Bonilla et al. are right to push for a broader definition, but the studies they found show how far the field still is from practicing it.
Tools for Inclusive Education: The Full Catalogue
The review documents an impressive range of tools. Mobile devices and tablets are the workhorses, appearing across nearly every category of inclusion. They’re portable, relatively affordable, and flexible enough to run different accessibility software depending on the student’s needs.
Physical and sensory accessibility tools like electronic Braille displays, adapted keyboards, robotic gloves, and 3D printers for tactile learning materials have been around for years. Navas-Bonilla et al. show that researchers are now studying them more systematically. Software tools like screen readers, text-to-speech programs (JAWS, Natural Reader), sign language translation apps, and voice writing tools form a second layer focused on removing barriers to content access.
Augmented and virtual reality appeared in 26 studies as collaboration and interaction tools. Educational robotics showed up in 16, often with students on the autism spectrum or those with intellectual disabilities. The three most common characteristics of these inclusive technologies were multimedia content (41 studies), accessibility (36 studies), and interactive content (32 studies). Ease of use, multisensory design, and adaptability appeared less frequently but were flagged as critical for real-world adoption.
I find the catalogue useful as a reference, but it raises a question the paper doesn’t fully answer: which of these tools actually work, and for whom? The review identifies what exists. It tells us less about what succeeds.
The Teacher Training Problem (Again)
The most persistent finding across the 159 studies is about teachers, not technology. Navas-Bonilla et al. report that insufficient training limits teachers’ ability to integrate inclusive technologies into their practice. The pattern of giving teachers tools without building their capacity repeats across multiple countries and contexts.
This tracks with everything I’ve covered on this blog. Celik’s Intelligent-TPACK framework (2023) made the case that teachers need a specific knowledge base for working with AI, not just general digital literacy. The same logic applies here. A teacher who receives an adapted keyboard or an AR application without understanding how to embed it in their pedagogy won’t use it well. The technology becomes decoration.
The digital divide compounds the problem. Students in rural or low-income areas lack basic infrastructure, let alone specialized assistive technology. Navas-Bonilla et al. found that unequal access to hardware, software, and connectivity amplifies the very inequalities inclusive technology is supposed to address. That’s a contradiction the field hasn’t resolved, and this review documents it without offering a clear path forward.
What the Review Misses
The most notable gap is generative AI. This paper was published in February 2025, reviewing literature up to 2024, yet GenAI tools barely register in the analysis. The review covers chatbots and intelligent tutoring systems in passing, but the explosion of tools like ChatGPT, which reshape how students with learning difficulties can access information, draft text, and process complex material, goes largely unaddressed.
That’s not entirely the authors’ fault. The studies they reviewed were published before GenAI became mainstream. But it does mean the review is already a partial snapshot. A teacher reading this paper in 2026 faces a technology picture that looks fundamentally different from what it describes.
GenAI is now embedded in mainstream productivity tools. Students with dyslexia, ADHD, and language barriers are already using it daily, often without any institutional guidance. The inclusion conversation has moved, and this review captures where it was, not where it is.
The authors also acknowledge their reliance on Scopus alone, which may have excluded relevant work from other databases. And the review gives a broad view across disability types without drilling deeply into any single one. That breadth is both a strength and a limitation.
Toward Inclusive AI Pedagogy
The strongest claim in the paper is about what inclusion actually requires. Navas-Bonilla et al. write that successful integration of technology “goes beyond simply adapting to individual needs” and “represents a decisive step toward building an educational system that recognizes diversity as a central value.” I agree with that completely. Inclusion is an architecture.
What I’d add, and what this review points toward without fully developing, is that inclusion in 2026 requires AI pedagogy. The tools have changed. AR headsets and Braille printers still matter, but a student with a learning disability now has access to a conversational AI that can explain a concept seventeen different ways in thirty seconds. Nieminen and Eaton’s work (2024) on how accommodation policies get framed as cheating reminds us that access tools only work when institutions treat them as legitimate. The same principle applies to GenAI as an inclusive technology: it only works if teachers and institutions design around it with intention.
The review by Navas-Bonilla et al. is a solid map of where inclusive education technology has been. The work ahead is building the pedagogy that makes these tools count.
References
- Celik, I. (2023). Towards Intelligent-TPACK: An empirical study on teachers’ professional knowledge to ethically integrate artificial intelligence (AI)-based tools into education. Computers in Human Behavior, 138, 107468. https://doi.org/10.1016/j.chb.2022.107468
- Linsenmayer, E. (2025). Leveraging artificial intelligence to support students with special education needs (OECD Artificial Intelligence Papers No. 46). OECD Publishing. https://doi.org/10.1787/edu/wkp(2025)12https://medkharbach.com/ai-in-special-education
- Navas-Bonilla, C. del R., Guerra-Arango, J. A., Oviedo-Guado, D. A., & Murillo-Noriega, D. E. (2025). Inclusive education through technology: A systematic review of types, tools and characteristics. Frontiers in Education, 10, 1527851. https://doi.org/10.3389/feduc.2025.1527851
- Nieminen, J. H., & Eaton, S. E. (2024). Are assessment accommodations cheating? A critical policy analysis. Assessment & Evaluation in Higher Education, 49(7), 978–993. https://doi.org/10.1080/02602938.2023.2259632