AI Governance for healthcare: A global perspective

AI Governance approaches vary greatly between regions. What does this mean for healthcare organizations worldwide and in the EU in particular?

AI is inherently more complex and unpredictable than other technologies, which leads to numerous ethical and legal concerns, necessitating new policies for the development and use of AI [1, 2]. These regulatory frames can either hinder innovation or foster more profound and safe integration due to clear guidelines and standardisation [3]. The World Health Organization (WHO) published ethical principles for the use of AI in health as a foundation for governments, developers, companies and international organizations to guide ethical adoption [4]. These principles, for example “Ensuring inclusiveness and equity”, are often described as “high level guidelines”[5] that leave room for interpretation. Consequently, approaches to AI governance in healthcare vary across regional and cultural contexts [1]. This science digest sheds light on the different policy-context of the EU, USA and China, that shape the environment for healthcare organizations which are in the midst of implementing AI.

Different approaches on AI Governance

In a recent study Shangrui Wang, Yuanmeng Zhang, Yiming Xiao and Zheng Liang analyzed 139 policy texts on AI from the European Union (EU), the United States (US), and China concerning AI Governance [1]. The documents, published between 2016 to 2023 were extracted from authoritative databases from the respective regions. The policies were then analyzed using the method of structural topic modeling (STM). This method allows not only to identify topics in texts but also to analyze relationship between these topics and with other variables like document type or publication date [6].

Differences and common ground in global policies

The authors identified 13 primary topics in the AI policy frameworks and cluster them in three main themes. The three regional regulatory contexts can be assigned to the three topics based on their governance priorities (Figure 1).

Figure 1: AI Governance Themes linked to China, US, and the EU (topics are not exclusively relevant in the respective areas)

China “Research and Application”

In China, the central government dominates the strategic direction of AI development, but allows for local adaptation and iterative policy formulation. The AI industry and the scientific community operate under broad policies with a great deal of autonomy in AI development. The training of talents is pursued by multiple initiatives to include AI skills in educational frameworks. Scientific advancements and broad application in the industry are also prioritized and receive comprehensive funding.

US: “Government Role”

Interestingly, in the US, often perceived as an ideal type for liberal markets, AI policies strongly focus on the government role. AI development is supported through national strategies, subsidies or investments and specialized agencies. To maintain a leadership position in global AI development, and to reinforce national security, the government promotes access to foreign markets to stimulate industrial growth and shape international standards. On the other hand, policies increasingly rely on technological restrictions to constrain competitors and protect critical innovations.

EU: “Social Impact”

The EU policies clearly show a more cautious AI Governance approach, focusing on the reinforcement of human rights and public welfare. As a political entity with supranational and intergovernmental elements, the EU aims to harmonize European markets and thereby establish itself as an “ethical leader”. This so-called “Brussels Effect” allows the EU to influence global regulatory standards to align with European values, which strengthens global competitiveness and, in turn, legitimizes the EU itself.

Similarities

While the study identifies strong differences in the direction of AI governance policies between the different regions, similarities also persist. While the EU has always focused on scientific insights, human rights and institutional systems, an increasing focus on these aspects can also be observed in other regions. This may be due to growing demands for improved regulation from the scientific community, international dependencies in the field of AI, and the experience of negative events such as data loss or market turmoil.

Significance for organisational AI Governance for healthcare

The study reveals the different regulatory contexts across global regions, but it also identifies common trends. These overarching AI policies directly shape the possibilities for developing and deploying AI solutions in healthcare. A key task for hospitals is to translate often vague regulatory and ethical requirements into concrete organizational structures and processes. In the EU context, AI Governance policies aim to coordinate the efforts of the member states. Particularly noteworthy is the AI Act [7], the first binding regulation worldwide for the development, deployment, and use of AI technology [8], which classifies many healthcare applications as high-risk systems. In this context, leaders of healthcare organizations could pursue closer cross-country cooperation, establish shared learning processes, and provide Brussels with consolidated feedback on their joint experiences and needs. The study also emphasizes the crucial role of science in a field marked by uncertainty and limited experience. While robust scientific evidence on organizational AI Governance in healthcare is still lacking, first case studies and implementation experiences are beginning to emerge (including on the ATLAS). To stay one step ahead in effective organizational governance, hospital management can actively monitor current scientific findings.

Bjoern Gostmann

Literature

1. Wang, S., et al., Artificial intelligence policy frameworks in China, the European Union and the United States: An analysis based on structure topic model. Technological Forecasting and Social Change, 2025. 212: p. 123971.

1. Mennella, C., et al., Ethical and regulatory challenges of AI technologies in healthcare: A narrative review. Heliyon, 2024. 10(4).

1. Leenen, J.P.L., et al., Exploring the complex nature of implementation of Artificial intelligence in clinical practice: an interview study with healthcare professionals, researchers and Policy and Governance Experts. PLOS Digital Health, 2025. 4(5 May).

1. Ethics and governance of artificial intelligence for health: WHO guidance., W.H. Organization, Editor. 2021, World Health Organization: Geneva.

1. Saenz, A.D., et al., Establishing responsible use of AI guidelines: a comprehensive case study for healthcare institutions. npj Digital Medicine, 2024. 7(1): p. 348.

1. Hollibaugh, G.E., The Use of Text as Data Methods in Public Administration: A Review and an Application to Agency Priorities. Journal of Public Administration Research and Theory, 2018. 29(3): p. 474-490.

1. REGULATION (EU) 2024/1689 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 13 June 2024 laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act), E. Union, Editor. 2024.

1. Schuett, J., Risk Management in the Artificial Intelligence Act. European Journal of Risk Regulation, 2024. 15(2): p. 367-385.