WHO elevates AI governance as a health-system priority in the Eastern Mediterranean
On the final day of the Seventy-third session of the World Health Organization Regional Committee for the Eastern Mediterranean, senior health leaders placed responsible AI governance squarely inside the public health agenda. The technical dialogue in Cairo focused on Artificial Intelligence for Health in the WHO Eastern Mediterranean Region: Priorities and Considerations, with policy, investment, and regulation positioned as the conditions that will determine whether AI strengthens care or widens existing gaps.
The message was direct. AI is already moving into health systems across the region, and the challenge is no longer whether it will be used, but how it will be governed.
Governance before scale in fragile and unequal health systems
Keynote remarks from H.E. Dr Musab Nazzal Al-Ali, Minister of Health of the Syrian Arab Republic, and H.E. Dr Rakan Nassereldine, Minister of Health of Lebanon, framed advanced technologies as tools that could support care delivery in fragile and complex settings. Those settings are precisely where weak oversight can do the most damage. When health systems face workforce shortages, data fragmentation, and uneven access to services, AI can either reduce friction at the point of care or harden existing inequities if deployed without guardrails.
Dr Arash Rashidian, Director a.i., Health Systems and Life Course at the WHO Regional Office for the Eastern Mediterranean, captured the speed of the transition by noting that AI is no longer a distant possibility and is already shaping health care and health systems across contexts ranging from capital cities to remote rural and nomadic areas. The scope of that statement matters because it places AI outside the narrow frame of pilot projects. Health ministries are being asked to manage a technology that is already entering clinical workflows, data systems, and service delivery models.
Policy, not software alone, became the central issue.
Data infrastructure and interoperability came before algorithms
Panelists focused on the practical requirements for responsible implementation: partnerships and investments, the data infrastructure needed to support AI, and the use of AI at the point of care and within clinical workflows. The emphasis on infrastructure is clinically and operationally important. AI tools depend on clean, interoperable, well-governed data, and health systems that cannot exchange information reliably will struggle to use these tools safely or fairly.
Dr Adham Ismail, Director of Programme Management and Officer-in-Charge at the WHO Regional Office for the Eastern Mediterranean, argued for investing in interoperability before scaling systems and in data before algorithms. In health terms, that sequence is hard to dispute. A model trained on incomplete, biased, or disconnected data can only amplify the weaknesses already embedded in a system. Without interoperable records, governance frameworks, and defined accountability, even technically sophisticated tools may produce clinically misleading outputs or operational confusion.
The discussion pointed to a familiar health systems truth: technology inherits the quality of the system around it. If the inputs are unreliable, the outputs become hard to trust in patient-facing settings where precision and accountability are nonnegotiable.
Minimum requirements for safe use in clinical workflows
The dialogue aimed to build consensus around policy and regulatory frameworks, define minimum requirements for Member States to establish AI governance for health, obtain commitments to create a regional technical advisory group for AI, and review actionable recommendations for safe, secure, and trustworthy deployment. Those goals place human oversight at the center of implementation rather than treating it as a symbolic addition.
Transparency, accountability, and robust human oversight were presented as the foundation of trustworthy AI in health. In clinical workflows, those concepts are not abstract governance language. They determine whether clinicians know how an AI recommendation was generated, whether responsibility is clearly assigned when a system fails, and whether the tool remains subordinate to medical judgment instead of displacing it.
A regional technical advisory group would give countries a shared mechanism for comparing policies, aligning standards, and responding to common problems. The Eastern Mediterranean includes health systems with very different capacities, but the regulatory questions are similar: what minimum data protections are required, how should AI be evaluated before use, who monitors performance after deployment, and what safeguards are necessary when tools are introduced into primary care, specialty services, or remote settings.
Regional collaboration and sector partnerships as the next step
Dr Anas Doukkali, President of the Centre d’Innovation et e-Santé at the University of Rabat in Morocco, Dr Ali Jafarian, Senior Advisor and Deputy Minister of Health and Medical Education in the Islamic Republic of Iran, and Dr Anderson Stanciole, Director of Strategic Performance at the Global Institute for Disease Elimination in the United Arab Emirates, contributed to a discussion that linked policy to implementation. The range of institutions represented at the session reflected the reality that AI governance cannot sit only inside ministries of health. Academic institutions, telecommunications partners, and health authorities all have roles in shaping whether digital systems remain usable, secure, and responsive to public needs.
Expected outcomes of the side event included practical insights on the policies and regulatory framework needed to support AI governance in the region, closer collaboration between the health sector, academic institutions, and the telecommunications sector, and identification of areas for action to promote prudent use of AI in health care. Those outcomes signal a move toward implementation rather than broad principle. Health systems do not need generic enthusiasm for AI; they need rules, infrastructure, and stewardship that can survive contact with real clinics and real patients.
The closing remarks from Dr Ismail centered the political and moral stakes. The objective, he said, is to ensure that AI accelerates progress toward health for all and the health-related Sustainable Development Goals, so that every country in the region can benefit. He also warned that the outcome will depend on the choices made now. In public health terms, that is the decisive issue: the technology is arriving, but whether it improves care will depend on governance that is strong enough to keep pace with deployment, adoption, and risk.
The session ended with a commitment from panelists to advance regional cooperation on shared AI challenges and to shape a future aligned with the region’s health priorities and values. The remaining work is concrete rather than theoretical: build interoperable systems, define minimum standards, assign responsibility, and make sure AI entering health services remains safe, secure, and trustworthy under human oversight.