Artificial intelligence (AI)-assisted technologies, such as decision support and monitoring systems, hold the potential to significantly improve efficiency and quality of care in the health sector. However, given the impact that such technologies can also have on work requirements and the moral agency of healthcare personnel, it is imperative - from an occupational safety and health perspective - to incorporate established criteria for human-centered work design and ethical design criteria into technology development throughout the entire life cycle. Existing AI guidelines and regulations such as the EU's AI Act address this imperative; however, suitable approaches for effectively integrating corresponding criteria into risk assessment and compliance processes are still lacking. This article presents the methodological approach and key findings from two multi-stakeholder dialogues, which identify starting points for the human-centered development of AI-assisted healthcare technologies.

This article is published in "TATuP - Journal for Technology Assessment in Theory and Practice" (2025).