Abstract

Despite the rapid advances in artificial intelligence (AI), its integration into healthcare practice remains limited. This study uses a socio-technical perspective to investigate factors that enable or hinder the implementation of AI in healthcare organisations. By exploring key prerequisites and challenges, the study aims to provide insight into how sustainable and meaningful implementation can be achieved in practice, particularly through the interplay between technical systems and human experience.

The study applies the Consolidated Framework for Implementation Research (CFIR) alongside complementary theories to analyse how AI is interpreted and integrated into clinical work. A theoretical model was developed to capture implementation dynamics from technical and individual perspectives. Empirical data were collected through sixteen semi-structured interviews with stakeholders across four professional domains (healthcare personnel, developers, strategists, and researchers).

The findings highlight that successful implementation depends not only on technology, but also on the alignment of AI with local values and practices. User involvement, clear responsibilities and interprofessional communication were identified as critical enablers. Informal change agents also played a central role in translating innovation into meaningful use. The study deepens our understanding of how innovations are either embraced or resisted in complex systems by gathering experiences of the implementation of AI in the Swedish healthcare system.