A Construction-Specific Framework Addressing Explainability, Liability, and Ethical Boundaries of AI Use in Civil Engineering

The Architecture, Engineering, and Construction (AEC) sector is increasingly adopting Artificial Intelligence (AI) technologies to enhance efficiency, decision-making, and operational performance across infrastructure and construction workflows. The objective of this study is to examine the challenges of implementing AI in civil and construction engineering, as associated with governance, ethics, and human interaction. The methodology integrates a structured conceptual framework developed through literature synthesis and exploratory survey analysis to assess the technical, governance, ethical, and sociotechnical implications of AI deployment in civil and construction engineering. The analysis focuses on key AI applications, including infrastructure inspection, bridge crack detection, construction safety monitoring, and predictive planning for maintenance and scheduling, while addressing critical dimensions such as explainability, liability, bias, ethical governance, and human factors. The results indicate that although AI-driven models can improve productivity and decision support, challenges related to transparency, accountability, data privacy, and overreliance on automated systems remain prevalent. The study concludes that the successful deployment of AI in civil and construction engineering requires rigorous validation practices, explainable and trustworthy system design, and sustained human-in-the-loop oversight to ensure responsible, reliable, and scalable integration into real-world engineering applications.