WHO - AI Governance Framework

Overview

The World Health Organization (WHO) serves as the leading global authority for health, extending its mandate to the ethical and responsible governance of artificial intelligence (AI) in healthcare. Recognizing AI's transformative potential to revolutionize diagnostics, treatment, and public health interventions, WHO simultaneously addresses the inherent risks associated with these technologies. Its comprehensive guidance aims to ensure that AI tools are developed and deployed in a manner that maximizes public health benefits, promotes global health equity, and rigorously safeguards human rights and dignity. This involves navigating complex and evolving challenges related to data protection, algorithmic bias, the potential for exacerbating existing health inequalities, and the digital divide, particularly in low- and middle-income countries (LMICs) where resources and infrastructure may be limited. WHO's work emphasizes a human-centered approach, ensuring that technology serves humanity's best interests rather than the reverse, and that the benefits of AI in health are accessible to all, not just a privileged few. The organization actively fosters international dialogue and collaboration to build a shared understanding and common standards for AI in health.

AI Governance Mandate

WHO's engagement in AI governance is deeply rooted in its overarching mission to promote health for all people, everywhere. The formal mandate for this work was established through the 2018 World Health Assembly resolution on Digital Health, which recognized the growing importance of digital technologies, including AI, in achieving health-related Sustainable Development Goals. This was further elaborated and solidified in the 'Global Strategy on Digital Health 2020–2025', which explicitly identifies artificial intelligence as a critical component within broader digital health initiatives. These foundational documents underscore the imperative for robust governance frameworks to ensure that AI applications are not only innovative and effective but also human-centered, ethical, secure, and contribute meaningfully to universal health coverage. WHO consistently stresses that ethical considerations, human rights, and principles of equity must be paramount and central to every stage of the design, development, deployment, and evaluation of AI technologies for health, preventing harm and fostering trust in these powerful tools.

Key AI Instruments

WHO has proactively developed several critical instruments and initiatives to guide and shape AI governance in the health sector globally:

• Ethics and governance of artificial intelligence for health (2021): This landmark report, developed through extensive and inclusive consultations with a diverse range of global experts, provides a foundational framework. It identifies six core ethical principles that must underpin the development and use of AI in health: protecting human autonomy (ensuring individuals retain control over their health decisions), promoting human well-being and public interest (prioritizing health outcomes and societal good), ensuring transparency and intelligibility (making AI systems understandable and their decisions explainable), fostering responsibility and accountability (assigning clear responsibility for AI outcomes), ensuring inclusiveness and equity (designing AI to benefit all, especially vulnerable populations), and promoting responsive and sustainable AI (ensuring systems are adaptable and environmentally sound). This report serves as a crucial reference for governments, technology developers, healthcare providers, and civil society organizations seeking to adopt ethical approaches to AI in health.
• Guidance on the ethics and governance of large multi-modal models (LMMs) (2024): In direct response to the rapid emergence and widespread adoption of generative AI technologies, particularly large multi-modal models (LMMs) like ChatGPT, WHO issued comprehensive guidance containing over 40 specific recommendations. These recommendations are tailored for governments, technology companies, and healthcare providers. The guidance aims to ensure the appropriate, safe, and responsible use of LMMs in public health and medical settings, addressing their unique ethical and governance challenges, such as the potential for generating misinformation, perpetuating biases, or creating 'hallucinations' that could lead to clinical errors. It emphasizes the need for rigorous testing, validation, and continuous monitoring of LMMs.
• ITU-WHO Focus Group on Artificial Intelligence for Health (FG-AI4H): Established in 2018 in partnership with the International Telecommunication Union (ITU), this collaborative focus group plays a vital role in operationalizing AI governance. It brings together an international network of experts from diverse backgrounds to work towards establishing a standardized assessment framework for the rigorous evaluation of AI-based methods used in health, particularly for critical applications like diagnosis, triage, or treatment decisions. The group develops technical specifications, guidelines, and benchmarking processes to ensure that AI solutions are robust, reliable, and safe for clinical use, fostering interoperability and trust across different systems and regions.

Implementation Mechanisms

WHO actively promotes the practical implementation of its AI governance guidance through a variety of strategic mechanisms designed to build capacity and foster collaboration at national and international levels. This includes organizing targeted capacity building workshops for key stakeholders, such as digital health focal points from Ministries of Health and Information Communication Technology, as well as WHO's own digital health focal points across its regional and country offices. These workshops are crucial for identifying country-specific actions for the responsible integration of AI in health, documenting best practices and lessons learned, and fostering open dialogue and collaboration among diverse stakeholders. Furthermore, WHO provides direct technical support and expert guidance to member states, assisting them in developing, strengthening, and aligning their national digital health strategies to responsibly integrate AI. This support ensures that national policies and regulatory frameworks are consistent with global ethical principles and best practices, facilitating a harmonized approach to AI in health worldwide.

Member Engagement

Member States are integral to the success of WHO's AI governance efforts, demonstrating strong commitment and active participation. They unanimously adopted the landmark resolution on Digital Health in 2018, which laid the groundwork for AI integration, and subsequently endorsed the comprehensive Global Strategy on Digital Health 2020–2025. This endorsement signifies a collective commitment to leveraging digital technologies, including AI, for health improvement. Member States are actively engaged in various consultative processes, technical workshops, and collaborative initiatives organized by WHO. This engagement is crucial for ensuring that global standards and ethical guidelines for AI are relevant, practical, and adaptable to diverse national contexts. WHO's approach emphasizes a collaborative, multi-stakeholder engagement model, recognizing that effective AI governance requires input and cooperation from governments, civil society, academia, the private sector, and affected communities at all stages of AI development and deployment, from policy formulation to practical implementation.

Current Focus Areas

Currently, WHO's focus areas in AI governance are multifaceted, addressing both the ethical challenges and practical risks associated with AI in health. A primary concern is the potential for biases encoded in training data, which can lead to discriminatory outcomes and exacerbate health inequities, particularly for marginalized populations. Robust data protection and privacy measures are also paramount, given the sensitive nature of health data and the increasing sophistication of AI systems. Cybersecurity vulnerabilities pose another significant threat, requiring continuous vigilance to protect patient information and system integrity. There is a strong emphasis on ensuring inclusiveness and equity, actively working to overcome the digital divide and ensuring that the benefits of AI reach underserved populations and low- and middle-income countries (LMICs) without creating new disparities. Furthermore, the responsible oversight and regulation of rapidly evolving AI technologies, including the imperative for mandatory post-release auditing and comprehensive impact assessments for large multi-modal models (LMMs), remain a critical priority to ensure ongoing safety, effectiveness, and ethical compliance throughout the AI lifecycle.

Future Directions

Looking ahead, WHO remains steadfast in its commitment to fostering global cooperation for the effective and equitable regulation of AI technologies in health. Future directions involve the ongoing development of nuanced guidance that keeps pace with technological advancements, ensuring that ethical frameworks remain relevant and robust. A key objective is to foster international alignment on AI standards and best practices, promoting interoperability and reducing fragmentation in global governance efforts. WHO will continue to support countries in building the necessary technical expertise, digital infrastructure, and regulatory capacity for the responsible adoption and deployment of AI. The overarching goal is to ensure that AI contributes significantly and sustainably to strengthening healthcare delivery systems worldwide, helping all countries make substantial progress towards achieving universal health coverage and the broader Sustainable Development Goals, ultimately improving health outcomes for everyone.

Resources

Resource	Description	Link
Ethics and governance of artificial intelligence for health: WHO guidance (2021)	First global report outlining six core ethical principles for AI in health.	https://www.who.int/publications/i/item/9789240029200
WHO releases AI ethics and governance guidance for large multi-modal models (2024)	News release on new guidance for LMMs in healthcare.	https://www.who.int/news/item/18-01-2024-who-releases-ai-ethics-and-governance-guidance-for-large-multi-modal-models
Global strategy on digital health 2020–2025	Strategic framework for promoting digital health, including AI, globally.	https://www.who.int/publications/i/item/9789240020924
ITU-WHO Focus Group on Artificial Intelligence for Health (FG-AI4H)	Information on the joint initiative for standardized AI assessment frameworks.	https://www.itu.int/en/ITU-T/focusgroups/ai4h/Pages/default.aspx