<i>desAIner</i> : Design, Implementation, and Evaluation of an AI Tool for Construction Hazard Prevention Through Design (CHPtD) Education

ABSTRACT Construction Hazard Prevention through Design (CHPtD) is widely recognized as a proactive strategy for reducing worker risk by addressing hazards during planning and design. However, its integration into architecture, engineering, and construction (AEC) education remains limited, leaving many students, who represent the next generation of industry professionals, underprepared to identify design‐related hazards and apply effective safety controls. If hazards remain unrecognized, they cannot be effectively addressed early in the project lifecycle. This study developed and evaluated desAIner , a domain‐specific, multimodal, human‐in‐the‐loop educational tool designed to support CHPtD learning through structured hazard recognition and mitigation planning grounded in the hierarchy of controls (HOC). Using a quasi‐experimental repeated‐measure counterbalanced design with construction management students, the study examined whether AI‐assisted support could strengthen students' ability to identify design‐related hazards and propose more effective prevention strategies. The findings indicate that desAIner improved students' CHPtD performance and supported more consistent application of prevention‐oriented reasoning across multiple construction scenarios. More broadly, the results suggest that carefully structured generative AI tools can help bridge persistent instructional gaps in CHPtD education by making design‐stage safety analysis more interactive and scalable within existing AEC curricula. This study contributes to the growing literature on AI in construction education by demonstrating how domain‐specific, human‐in‐the‐loop systems can support the development of safer design thinking among future construction professionals.