Refining collaborative AI literacy: A systematic review of human–AI collaboration competency and AI literacy measures

Abstract

With the widespread adoption of artificial intelligence (AI) and its growing role in solving complex problems, human–AI interaction has evolved into a new dimension, where AI is increasingly perceived as a collaborative partner. However, current AI literacy frameworks focus on technical knowledge and ethical use, often neglecting the ability to collaborate effectively with AI as a standalone construct. Conceptualized under the umbrella of AI literacy, the boundaries of human–AI collaboration competency need to be refined in order to create common ground for advancing research and measurement of this essential ability. This systematic review synthesizes 17 studies conducted in higher education settings to examine how human–AI collaboration competency is represented and measured within existing AI literacy scales. The findings reveal that human–AI collaboration is rarely explicitly assessed, let alone as a distinct construct. Some studies included a small number of items that could implicitly refer to collaboration competency or context, but most included no such representation. Drawing on these findings, we propose theoretical implications regarding contexts and factors that should be considered when developing new items, and measurement implications for three possible sub-factors of human–AI collaboration competency: coordination, communication, and metacognition. Overall, the findings provide a foundation for reconceptualizing human–AI collaboration competency within AI literacy, facilitating the development of measurement tools and instructional guidelines that can both assess and foster collaborative AI literacy.