Understanding AI adoption in healthcare organisations through interacting technological, organisational, and environmental conditions

Purpose Artificial intelligence (AI) is widely recognised as a transformative force in healthcare, yet its organisational adoption remains uneven and fragmented across contexts. Existing studies have identified technological, organisational, and environmental conditions influencing AI adoption, but these factors are often examined independently, resulting in inconsistent and inconclusive findings. This study addresses this gap by examining how organisational AI implementation and sustained-use decisions emerge through the interaction of these conditions rather than through isolated effects. Drawing on the Technology–Organisation–Environment (TOE) framework, the study aims to develop a more integrated and context-sensitive understanding of AI adoption in healthcare organisations operating across diverse institutional and regulatory environments. Design/methodology/approach The study adopts an inductive qualitative research design grounded in the TOE framework. Data were collected through semi-structured interviews with 29 healthcare experts actively involved in AI adoption across multiple national contexts. These participants represented a range of roles, including clinical leaders, technology specialists, and policy stakeholders, ensuring diverse perspectives on adoption processes. The data were analysed using a Gioia-inspired methodology, enabling systematic identification of first-order concepts, second-order themes, and aggregate dimensions. This approach facilitated the development of a process-oriented and configurational understanding of how technological, organisational, and environmental conditions interact to influence organisational AI implementation and sustained use. Findings The analysis identifies six key dimensions influencing AI adoption. Environmental conditions include risk allocation ambiguity and national digital scaffolding. Organisational conditions comprise economic alignment and capability orchestration. Technological conditions involve epistemic transparency and data robustness. A pattern-oriented synthesis reveals that AI adoption emerges through recurring interaction patterns across these dimensions. The findings demonstrate that adoption is not driven by isolated factors such as performance expectations, organisational readiness, or regulatory pressure alone. Instead, technological attributes become consequential only when aligned with organisational capabilities and regulatory accountability, highlighting the fundamentally interactional nature of AI adoption in healthcare contexts. Research limitations/implications This study has several limitations. First, the qualitative design and sample of 29 experts support analytical rather than statistical generalisation across diverse healthcare settings. Second, the findings rely on participant perceptions, which may introduce subjectivity and potential bias. Third, although the multi-country sample adds richness, it may obscure important country-specific regulatory and infrastructural differences. Fourth, the cross-sectional design does not capture how AI adoption evolves over time. Finally, the study focuses primarily on expert-level perspectives, with limited representation of frontline healthcare professionals. Future research should adopt longitudinal and mixed-method approaches to examine these interaction patterns across larger and more diverse samples and explore context-specific variations in greater depth. Practical implications The findings offer important implications for healthcare managers, policymakers, and technology providers. Organisations should move beyond focusing on individual adoption drivers and instead prioritise alignment across technological capabilities, organisational readiness, and regulatory frameworks. Policymakers should address risk allocation ambiguity and strengthen national digital infrastructures to support AI integration. Healthcare leaders should invest in capability orchestration, ensuring that technical expertise, clinical knowledge, and organisational processes are effectively integrated. Technology developers should enhance epistemic transparency and data robustness to build trust and usability. Overall, a coordinated and system-level approach is essential for enabling effective and sustainable AI adoption in healthcare organisations. Social implications AI adoption in healthcare has significant implications for patient outcomes, equity, and system sustainability. By highlighting the importance of interactions across technological, organisational, and environmental conditions, this study provides insights into conditions that may support more responsible and effective organisational AI implementation. Greater attention to these interacting conditions may help healthcare organisations realise potential benefits for clinical practice and patient care while managing unintended risks associated with misaligned systems. Originality/value The originality of this study lies in its interactional and process-oriented perspective, moving beyond static models of adoption. It offers a novel explanation for fragmented adoption patterns and provides a comprehensive framework for understanding AI integration in complex healthcare settings.