As the construction industry continues to embrace digitalization, machine learning is emerging as a powerful tool to improve efficiency and enhance decision-making throughout the project lifecycle. This study presents an exploration of the potential of machine learning in the construction industry, focusing on reducing accident risks through decision support. Through a combination of workshops and analysis of machine learning applications in other sectors, this study provides insights into the opportunities and challenges associated with machine learning in construction. Results suggest that machine learning tools can enhance information-gathering, visualization of trends, prediction of outcomes, and evaluation of effectiveness. However, challenges such as increased complexity, criticality, and lack of trust in machine learning must be addressed. The study recommends developing a theoretic safety model for machine learning tools, focusing on finding the correct parameters and addressing challenges associated with machine learning. Overall, the study concludes that adopting machine learning can benefit construction, but it is essential to consider these challenges carefully.