This large-scale international collaboration, coordinated by the University of Edinburgh, the University of Oxford, and Zhejiang University, represents the first structured global priority-setting exercise focused specifically on the application of AI and data science in global health.
Using the established CHNRI (Child Health and Nutrition Research Initiative) methodology, 75 international experts generated 155 research questions, which were subsequently evaluated by 51 experts across five key criteria: feasibility, potential to reduce disease burden, paradigm-shift potential, likelihood of implementation, and impact on equity.
The study identifies several urgent global priorities, including:
- epidemic preparedness and early warning systems,
- diagnostic innovation (including tuberculosis, malaria, and COVID-19),
- health system strengthening and resource optimization, and
- chronic disease management, particularly in low- and middle-income country (LMIC) contexts.
A particularly novel contribution of the research is the empirical comparison between expert-defined priorities and outputs generated by leading large language models. This provides early insights into how generative AI may complement—though not replace—structured expert judgement in research governance and priority setting.
Contributing authors from IEDC and BILDAI include Igor Rudan, Stjepan Orešković, Mili Sanwalka, and Brane Kalpič. Their participation reflects the institutions’ ongoing commitment to responsible AI, leadership in digital transformation, and evidence-based approaches to global governance.
IEDC and BILDAI extend their sincere appreciation to Professor Igor Rudan, Professor Aziz Sheikh, and Professor Peige Song for their intellectual leadership and coordination of this significant global initiative.
For IEDC and BILDAI, this publication reinforces a clear direction: the development and application of artificial intelligence—whether in health, policy, or industry—must be grounded in evidence, guided by principles of equity, and aligned with long-term societal value.
The open-access article is available here:
https://lnkd.in/d77ehhit