What happened
Fatiha Tali-Otmani argues that the systems now entering research and teaching are not neutral knowledge pipes. They inherit Anglophone, Western-centric training distributions, then convert those distributions into answers, rubrics, summaries, and institutional defaults. The disability case is the sharp edge: when disabled people are reduced to stereotypes or left out of design, AI can marginalize their knowledge twice, first in the data and then in the workflow.
That matters because epistemology is becoming an infrastructure question. The issue is not only whether an answer is correct. It is whose ways of knowing become easy to retrieve, legitimate to cite, and cheap to operationalize.
The practical test for universities and builders is simple: if the system cannot show where its knowledge world is thin, it should not be treated as a neutral authority.
Source
Reported by Generative artificial intelligence and the marginalization of minoritized knowledges in higher education: the case of disability via arxiv.org, published May 26, 2026.