Disability, Data, and Design: Toward a Critical Framework for Evaluating AI in Inclusive Education

In today’s digital landscape, artificial intelligence (AI) is significantly influencing higher education by driving adaptive learning platforms, automated assessment, predictive analytics, and generative AI tools.From adaptive learning systems to predictive analytics and automated assessment, AI is transforming higher education in today’s digital era. The developments bring up some significant issues relating to the impact these will have on disabled learners. Though the issues of equity and inclusion are common to both fields, AI in Education (AIED) and critical disability studies have developed separately for the most part. This paper presents a critical evaluative approach to evaluate the impacts of AI systems in education on reinforcing, challenging or changing ableist structures. The framework centres on four aspects: (1) the range of assumptions of a normed nondisabled learner that may be found in system design and training data; (2) whose ability can be facilitated by AI adaptations; (3) the politics of disclosure, personalisation and surveillance; and (4) whose knowledge and perspectives inform system development. It summarizes existing research on both sides, proposes theoretical links, and illustrates its applicability in three educational AI scenarios, providing a basis for empirical and critical research.