The digital transformation of education necessitates innovative solutions that are both technologically advanced and equitable. This study addresses the compounded digital exclusion faced by women with disabilities by developing and evaluating IncluLearn AI, a novel adaptive AI-driven tutoring system. Grounded in a user-centered framework that integrates Universal Design for Learning (UDL), Participatory Design (PD), and Explainable AI (XAI) principles, the system aims to deliver personalized, accessible, and transparent digital literacy training. Using a sequential exploratory mixed-methods design, we conducted co-design workshops with 45 women with disabilities in Jordan, followed by a 12-week quasi-experimental field study comparing an intervention group ( n = 30) using IncluLearn AI with a control group ( n = 15) using a traditional self-paced digital course. Quantitative results revealed that the intervention group achieved significantly greater improvements in digital literacy performance F (2,86) = 18.34, p .001, η 2 = .30 and digital learning self-efficacy F (2,86) = 12.77, p .001, η 2 = .23 compared to the control group. The system received high usability ratings (SUS M = 82.4) and demonstrated strong user acceptance (Perceived Usefulness M = 4.5/5). Qualitative analysis identified four key themes: Agency Through Customization, Trust via Transparency, Competence from Adaptive Challenge, and Inclusion by Design. This research provides empirical evidence that AI-driven digital learning tools, when intentionally designed with inclusion, participation, and transparency at their core, can effectively foster digital skill acquisition and empowerment among marginalized learners. The study offers a replicable framework and evidence-based guidelines for creating equitable AI-enhanced learning environments within digital education.
Maberah et al. (Mon,) studied this question.