Abstract
This article examines the role of artificial intelligence (AI) in monitoring human factors and improving occupational safety in smart cities. The aim of the paper is to identify directions for AI use in occupational health and safety systems and assess ethical and legal challenges. Smart cities leverage integrated data, IoT, and machine learning to predictively respond to threats, enabling a shift from a reactive to a preventive model. The authors distinguish four main areas of AI application: monitoring employee physiological status, automatic recognition of personal protective equipment (PPE), predictive risk models, and occupational health and safety training in a VR/AR environment. CNN and LSTM algorithms enable high-accuracy fatigue and stress detection, while systems based on YOLOv8 and Mask R-CNN automatically identify safety violations. The article presents examples of implementations in Poland (SmartCity Poznań 2030+, Urban Safety AI, PGE SafetyNet) and the EU, confirming the effectiveness of AI in reducing accidents. The importance of the AI Act and GDPR regulations regarding human oversight, algorithm transparency, and biometric data protection is emphasized. Attention was also drawn to the ethical dilemmas associated with automated supervision and the need to maintain the “human-in-the-loop” principle. The conclusions indicate that AI is a key element of a modern occupational health and safety system, increasing work efficiency and safety. However, its success depends on a balance between technology and human responsibility.