How human-AI collaboration event strength shapes information security behavior: evidence from a University in Shanghai.

INTRODUCTION: As artificial intelligence is increasingly integrated into higher education learning contexts, university students receive technological support while also facing new information security risks. However, existing research has paid limited attention to how the event characteristics of human-AI collaboration are associated with university students' information security behavior and to the psychological processes underlying these relationships. Drawing on event system theory, this study develops a dual pathway model to examine how human-AI collaboration event strength, reflected in event novelty, event disruption, and event criticality, is associated with information security behavior among university students through security vigilance and cognitive overload.

METHODS: Data were collected from students at a university in Shanghai using a three-wave time-lagged survey design, yielding 351 valid responses.

RESULTS: The results show that event novelty, event disruption, and event criticality are positively associated with security vigilance and cognitive overload. Security vigilance is positively associated with information security behavior, whereas cognitive overload is negatively associated with information security behavior. Further analyses indicate that these event strength attributes generate positive indirect effects through security vigilance and negative indirect effects through cognitive overload.

DISCUSSION: This study explains information security behavior during university students' use of artificial intelligence from an event-strength perspective and provides a more nuanced understanding of the underlying psychological processes. The findings provide a theoretical basis for universities to refine guidelines for artificial intelligence use, strengthen information security education, and optimize the design of human-AI collaborative learning.