Abstract

Interdisciplinary professional work relies on relational coordination, carried out through shared workflows across occupational groups. When generative artificial intelligence (GenAI) enters these workflows, it produces communicative content that intervenes directly in that coordination, contesting the grounds on which professional authority and responsibility rest. Because these outputs keep evolving after deployment, sociotechnical principles – meant to make work efficient and meaningful for the people who do it – cannot be fixed at the design stage and instead require ongoing negotiation. These conditions call for new approaches. Drawing on Action Design Research, this study introduces participatory refinement of sociotechnical principles, examined in a hospital department introducing a GenAI-based virtual assistant, where doctors, nurses, and medical secretaries negotiated sociotechnical principles. The findings reveal three recalibrations of these sociotechnical principles: from displacement to substitution as the primary threat to wholeness, from ongoing negotiation to continuous governance in joint optimization, and from co-evolution to contextual anchoring in adaptability. These recalibrations resulted in six sociotechnical principles balancing efficiency, professional judgment, and relational care. The study contributes theoretically by demonstrating why sociotechnical principles require recalibration, and methodologically by introducing participatory principle refinement for GenAI integration.

Download

Share

COinS