AI Tools for Supporting Students with ADHD in Higher Education in Gulf Cooperation Council (GCC) States: A Systematic Literature Review

Abstract

Although a considerable number of states in the Gulf Cooperation Council (GCC) and nationally stipulated AI strategies have a high prevalence of ADHD, there is no peer-reviewed data on AI-aided ADHD interventions in higher education in the GCC. This systematic review will answer three research questions namely: transferability of evidence to GCC settings; ethical, cultural, linguistic and privacy concerns to GCC deployment; priorities of a specific research agenda to the GCC (RQ3). This review, which is embedded in a 63-source corpus obtained based on systematic bilingual search (k = 0.84) was conducted in accordance with PRISMA 2020 with JBI critical appraisal. The transferability assessment matrix of GCC was designed on the technological, linguistic, cultural, and structural planes. The analysis of the ethical results was systematic using a concern-to-investigation gap analysis, and thematic synthesis was based on the Thomas and Harden (2008) structure. The lack of peer-reviewed studies in any GCC state was zero, which validates evidence vacuum of a systematic character. Transferability testing demonstrated high technological portability; low linguistic transferability, as the 63 tools were all English-only and Arabic NLP capacity remains a domain of underdevelopment; moderate cultural transferability, which involves the ability to port interventions based on core executive functions, but where family processes and stigma need to be culturally adjusted; and low structural transferability since GCC disability models were immature. Applicable ethical dimensions, five of which are identified as GCC-specific implications, and a proposal of an eight-priority research agenda was identified. The GCC evidence vacuum captures convergent influences of the stigma of ADHD, suppression of disclosure, the Western epistemic hegemony, and the infrastructure gap in the disability area. Reducing this gap necessitates Arabic AI co-design, disclosure-free support models, epidemiological baseline studies, as well as culturally modified implementation frameworks.