Teaching every learner: the AIMES Chair and the future of inclusive AI
New Research Chair · MIAI Cluster IA A new research chair is reframing what AI-powered education can look like — not as a text-based shortcut, but as a multimodal gateway to meaning for every learner.
Between 15 and 20 percent of pupils experience specific learning difficulties or neurodevelopmental disorders. Yet most educational tools — and most AI systems built to support them — still rely almost entirely on written text. For students who are deaf or hard of hearing, who use augmentative and alternative communication, or who live with cognitive or sensory differences, this creates a persistent and structural barrier to knowledge.
The AIMES Chair (Artificial Intelligence for Multimodal Educational Systems), funded within the MIAI Cluster IA and led by Saskia Mugnier (LIDILEM, Université Grenoble Alpes), sets out to change this. Rather than treating multimodality as an add-on, AIMES places it at the very centre of how AI can represent and transmit meaning.
Beyond text: a paradigm shift
The project builds on a well-established finding from linguistics and cognitive science: human beings learn multimodally — through language, gesture, image, space, and sign. AIMES translates this insight into AI systems capable of transforming educational content across these dimensions: text simplification, schematisation, visual narration, pictograms, and sign language mediation, all grounded in Universal Design for Learning principles.
AIMES is innovative not because it claims novelty, but because it changes how meaning, accessibility and AI are articulated in research.
The chair is structured around three interconnected work packages: a theoretical and analytical framework for multimodal meaning (WP1); the development of AI transformation pipelines (WP2); and empirical co-design studies with practitioners in real educational contexts, initially in the discipline of History (WP3).
A collaboration spanning French AI ecosystems
AIMES is a clear demonstration of what cross-ecosystem collaboration in French AI research can produce. The chair connects MIAI's expertise in AI and language technologies with the research community of Sorbonne Université through Mohamed Chetouani (ISIR), who contributes to the project's AI methodology, ethics, and governance. This link also ties AIMES into the scientific agenda of PostGenAI@Paris and SCAI — whose shared focus on evaluation, human-centred AI, and the societal implications of generative systems is directly relevant to the chair's ambitions.
The project carries a €400k budget over four years and will recruit a PhD student, a postdoctoral researcher, and a research engineer — all working toward an open-science model: open-access corpora, open-source tools, and documented methodological frameworks designed to outlast the funding period.
In a research landscape where AI for education often means chatbots and automated grading, AIMES offers something more principled: a scientific bet that the future of inclusive learning lies not in more text, but in richer, more human representations of meaning.