Adaptive interaction in intelligent augmented reality: design decision considerations and framework

Integrating Artificial Intelligence (AI) into Augmented Reality (AR) systems has evolved AR from a passive visualization tool into an intelligent medium for dynamic task execution. For this new role, Intelligent Augmented Reality (IAR) systems must adapt user interfaces (UIs) in real-time to users’ cognitive and physical states, task requirements, and environmental constraints, which raises foundational questions: what design decisions contribute to useful adaptive interaction? To address this question, this paper presents a systematic review of IAR systems through the lens of interaction design at the intersection of AR technologies , AI integration , and User Experience (UX) considerations . After screening 798 peer-reviewed works, 241 full-text records were assessed for eligibility. We reviewed 24 IAR published works based on their interaction features, including proactivity versus reactivity and communication styles and modalities, as well as their types of adaptation and control dynamics. Analyzing these features led to the development of a taxonomy for Levels of Automation and a top-down design framework for adaptive UIs in IAR Systems. Our findings highlight the effectiveness of assistive interaction and semi-autonomous IAR systems that achieve a balanced integration of adaptive assistance with prioritizing user agency while also revealing a noticeable gap in implicit understanding within current IAR implementations.