Artificial Intelligence and Public Health Data in Emergencies: A Critical Issue in the Middle East

Timely access to reliable public health data is a critical determinant of effective response to health emergencies, including disease outbreaks, climate-related health shocks, and humanitarian crises. Although artificial intelligence (AI) has been promoted for its potential to enhance outbreak prediction, situational awareness, and resource allocation, its effectiveness depends on the speed, quality, and accessibility of underlying data systems. Inequities in digital infrastructure, particularly in low- and middle-income settings, undermine prompt public health data management and in turn weaken emergency response capacity. With examples from epidemics, climate-related health emergencies, and conflict-affected settings in and beyond the Middle East, we examine how delays arise across public health data transmissions that include reliance on paper-based data collection, fragmented and incompatible databases, linguistic barriers in non-English data processing, dependence on externally hosted cloud infrastructure, and vulnerability to telecommunications shutdowns. Geographical concentration of data centers in high-income regions compounds these challenges by introducing latency into time-critical data processing and limiting local control over surveillance and analytics. Case studies from Yemen, Iraq, Sudan, and regional refugee responses illustrate how such delays erode surveillance data operational value, rendering early warning systems reactive rather than anticipatory. In contrast, settings with locally hosted, interoperable, and resilient digital infrastructure demonstrate the capacity for nearly immediate analysis and faster public health action. AI should be understood not as a standalone solution but as a downstream tool within a broader public health data ecosystem. Strengthening local digital infrastructure and governance is thus essential for timely, equitable, and effective public health emergency management.