Introduction
Artificial intelligence (AI) is increasingly integrated into workplaces, transforming business processes, task organization, and working conditions. From an occupational safety perspective, AI serves a dual role: as a workplace tool with potential risks and as a technology that can enhance risk assessments and promote workplace health.
This article explores the scientific foundations of AI’s application in occupational safety, based on research by Adolph and Westhoven (2025), focusing on its opportunities and challenges.
AI as a Workplace Tool
AI systems, as workplace tools, introduce changes to work processes but also present risks due to their probabilistic logic, which remains under active investigation (Messeri & Crockett, 2024). To ensure safe use, AI must comply with existing occupational safety regulations, including the EU Machinery Regulation and standards such as DIN EN ISO 9241, which emphasize ergonomics and user-friendliness (Adolph & Westhoven, 2025). These regulations require AI systems to be transparent, error-resilient, and human-centric to minimize risks to employees.
For instance, when AI is embedded in machinery or software applications, it must meet equipment and workplace safety standards. This includes evaluating human-system interactions and organizational factors, such as the allocation of responsibilities for overseeing AI functions. While specific AI standards are still in development, existing regulations provide a foundation for safe implementation.
AI in Risk Assessments and Workplace Health
AI offers significant potential to advance occupational safety by processing sensor data and natural language. These capabilities enable faster and more in-depth analysis of workplace conditions, particularly within risk assessments (Adolph & Westhoven, 2025). For example, AI can analyze sensor data to identify physical or psychological stressors, such as improper postures or high-intensity tasks, which can inform Employee Assistance Programs (EAPs) or optimize work processes.
Research from the Federal Institute for Occupational Safety and Health (BAuA) indicates that AI can streamline risk assessments, particularly in small and micro-enterprises where such processes are often neglected due to perceived administrative burdens (Westhoven & Adolph, 2022). By leveraging natural language processing models, AI can analyze textual workplace data, while time-series analysis from sensors enables continuous risk monitoring.
Challenges and Future Directions
Despite its potential, AI’s application in occupational safety faces challenges, including uncertainties in implementation and the lack of specific guidelines. Existing regulations, while applicable, are not fully tailored to AI’s unique characteristics, necessitating the development of new standards (Adolph & Westhoven, 2025). Additionally, organizational aspects, such as responsibility allocation and workplace social dynamics, must be considered to ensure a human-centered approach to AI integration.
Future research, such as BAuA’s ongoing projects, focuses on developing AI tools for automated risk assessments and integration with smart workplace systems. For instance, combining AI with sensor technologies could enable real-time interventions, such as adjusting workstations to reduce physical strain, which can be linked to EAPs to address psychological stress.
Conclusion
Artificial intelligence presents a dual opportunity in workplaces: as a tool requiring careful management and as a means to enhance occupational safety. By aligning with existing regulations and developing new standards, AI can improve safety and well-being through precise risk assessments and optimized work processes.
These technologies promise safer and healthier workplaces, provided they are implemented with a focus on human needs and regulatory requirements.
References:
• Adolph, L., & Westhoven, M. (2025). Artificial intelligence in the workplace and as a tool for occupational safety. ASU Arbeitsmed Sozialmed Umweltmed, 60(1).
https://doi.org/10.17147/asu-1-411953
• Messeri, L., & Crockett, M. J. (2024). Artificial intelligence in science. Cited in Adolph & Westhoven (2025).
• Westhoven, M., & Adolph, L. (2022). Research project on the application of artificial intelligence in occupational safety. Federal Institute for Occupational Safety and Health (BAuA).
