This systematic review investigates the application of Artificial Intelligence (AI) in reducing safety risks on construction sites. The research aims to synthesize existing studies on AI-driven safety risk management strategies and assess their effectiveness in enhancing safety efficiency, and sustainability within the construction environment. The review encompasses a variety of AI techniques, including machine learning, optimization, predictive analytics, and intelligent control systems, which are employed in safety management across different project functions. Additionally, the article examines the impact of AI on the application of Building Information Modeling (BIM) and the use of algorithms to improve AI model development. It examines the potential of AI across various domains, including Object Detection, Recognition, On-Site Monitoring and Active Warning; AI Supporting Human Decision-Making Processes; Safety Risk Management and Optimization; Safety Risk Assessment and Prediction; and Safety Risk Management and Early Warning Modeling. The findings highlight AI’s promising role in significantly mitigating potential hazards, with emerging applications that include real-time monitoring and control, as well as proactive, personalized operations for workers and equipment. Additionally, the study identifies research gaps and proposes future directions for exploration in this rapidly evolving field, emphasizing the importance of integrating AI solutions into comprehensive safety design and management strategies.