AI-POWERED ADAPTIVE USER INTERFACES FOR DIGITAL INCLUSION: A SYSTEMATIC REVIEW

Abstract

Around 763 million low-literate adults around the world face digital exclusion as they are unable to access important services in health, agriculture, education and e-commerce. These users tend to have requirements that are not met by traditional text-based user interfaces (UIs). In this systematic literature review (SLR), the state-of-the-art of AI powered adaptive UIs is explored, and the role of artificial intelligence (AI) in augmenting low-literate users' capabilities in terms of usability, accessibility, engagement, and autonomy is assessed. Using a structured search process, a total of 45 peer-reviewed studies published in the last eight years (2015–2023) were identified from the IEEE Xplore, ACM Digital Library, Scopus, and Web of Science databases related to AI-enhanced adaptive interfaces for low-literate and semi-literate users. Practical implementation, domain specificity and real-time adaptation were all emphasized for inclusion considerations, thereby ensuring actionable insights. Each of the studies spans a range of interface modalities from voice based to multi-modal to gesture-driven. The review indicates that voice-first AI interfaces with ASR and NLP are best suited for fully-illiterate users for low resource languages, and adaptive multi-modal interfaces with minimal text, culturally rich icons, and audio support for semi-literate users. In terms of numbers, 60% of studies relied on the voice-first approach and 40% on the adaptive multi-modal approach. Examples of innovations in domains include voice-control of healthcare workflows, camera-based agricultural advice, adaptive educational material, and icon-based e-commerce navigation. Based on these results, cross-domain AI-based design principles were extracted, highlighting modalities adaptation, context aware feedback, personalized navigation and adaptive learning strategies. These principles offer practical guidance for designers, developers, and policy makers to produce inclusive AI systems for digital tools to promote accessibility, usability, trust, and digital equity in relation to the United Nations Sustainable Development Goals (SDGs).