AI in one hand, inclusive education in the other: why mainstream teachers who use AI do not use it for inclusive education
DMDP-MFDP
- Besozzi, Roberta ORCID Université de Fribourg
- Menon, Anoop Eisengard AI
- 2026
Published in:
- International Journal of Inclusive Education. - Informa UK Limited. - 2026, p. 1-18
English
Mainstream Swiss teachers have adopted generative AI rapidly, yet rarely turn to it when facing inclusion difficulties – even though
the same tool that drafts their lesson plans could equally suggest strategies for a specific student. The gap between broad adoption of digital tools and their application to specific educational purposes is well documented. Still, existing explanations were
developed for technologies whose domains and functions were pre-specified by design. Generative AI has no such pre-specified
domains, raising an open question: why does the gap persist? This mixed-methods exploratory study addresses that question with a trilingual survey of 69 Swiss mainstream teachers and nine in-depth interviews with teachers and school leaders across three
cantons. We find a steep behavioural cascade: 74% of teachers use AI professionally, 18% extend it to inclusion-specific purposes, and only 4% name it as a coping strategy for inclusion difficulties. Three factors are associated with crossing to inclusion-specific use: breadth of AI engagement, materials creation as a bridge to student-specific adaptation, and professional distress. We theorise the pattern as a cognitive category boundary and organise the findings through the Boundary Model of AI in Inclusive Education.
the same tool that drafts their lesson plans could equally suggest strategies for a specific student. The gap between broad adoption of digital tools and their application to specific educational purposes is well documented. Still, existing explanations were
developed for technologies whose domains and functions were pre-specified by design. Generative AI has no such pre-specified
domains, raising an open question: why does the gap persist? This mixed-methods exploratory study addresses that question with a trilingual survey of 69 Swiss mainstream teachers and nine in-depth interviews with teachers and school leaders across three
cantons. We find a steep behavioural cascade: 74% of teachers use AI professionally, 18% extend it to inclusion-specific purposes, and only 4% name it as a coping strategy for inclusion difficulties. Three factors are associated with crossing to inclusion-specific use: breadth of AI engagement, materials creation as a bridge to student-specific adaptation, and professional distress. We theorise the pattern as a cognitive category boundary and organise the findings through the Boundary Model of AI in Inclusive Education.
- Faculty
- Faculté des sciences de l'éducation et de la formation
- Department
- Département de formation à l’enseignement
- Language
-
- English
- Classification
- Arts, Human and Social Science
- Other electronic version
- License
-
CC BY
- Open access status
- green
- Identifiers
-
- DOI 10.1080/13603116.2026.2715640
- ISSN 1360-3116
- Persistent URL
- https://folia.unifr.ch/unifr/documents/336503
Statistics
Document views: 34
File downloads:
- aiinonehandinclusiveeducationintheotherwhymainstreamteacherswhouseaidonotuseitforinclusiveeducation: 32