Evaluating the factors influencing artificial intelligence technology uptake in health and safety management within the Ghanaian construction industry

Purpose The construction industry in Ghana faces significant challenges in managing health and safety risks, leading to high rates of accidents and fatalities. Despite the potential of artificial intelligence (AI) technologies to improve health and safety management, their adoption in the Ghanaian construction industry remains limited. This paper aims to identify and evaluate key factors influencing the uptake of AI technologies in construction health and safety management within the Ghanaian industry. Design/methodology/approach The study adopts a rigorous two-step qualitative approach to identify a set of 17 variables. First, an extensive analysis of scholarly publications was conducted to compile an initial variable list. Secondly, a pilot survey involving both academic and industry professionals assisted in refining the identified variables. Subsequently, a questionnaire survey involving 219 Ghanaian construction professionals then collects quantitative assessments of each variable using the purposive sampling technique. Statistical modelling using factor analysis and fuzzy synthetic evaluation (FSE) was applied to process the survey data and determine the criticality of the factor categories. Findings The factor analysis yielded a three-factor solution underlying the 17 adoption variables: Extensive technological requirements and costs, resistance to change and AI adoption and uncertainty about AI outcomes and value. Subsequently, FSE confirmation showed the Extensive Technological Requirements category as the most critical, with specialized algorithmic demands, infrastructure limitations and expert support needs presenting major obstacles Ghanaian firms face in AI adoption. Originality/value This research contributes robust empirical evidence and novel factor-based statistical analysis to augment the theoretical discourse surrounding construction safety technology integration and change dynamics. The developed fuzzy quantitative methodology offers a model for assessing complex innovation adoption decisions in the face of uncertainty. The research addresses a gap in existing literature by providing a comprehensive assessment of the technological, organizational and environmental factors shaping AI adoption decisions and offering practical strategies for overcoming adoption barriers.