Impact of AI on public health emergency communication in Europe and Central Asia: A Delphi study

Abstract Background Artificial Intelligence (AI) holds the potential to fundamentally transform public health communication - presenting promising opportunities alongside profound challenges and risks. This two-wave Delphi study therefore invited practitioners and scholars to assess the potential impact of AI on the risk communication, community engagement, and infodemic management (RCCE-IM) of public health authorities in European and Central Asian countries. Methods In two successive online surveys, 54 experts - both RCCE-IM practitioners and scholars with expertise in (digital) health communication, risk communication, and AI - from 28 countries provided their assessment of the following analytical dimensions: (1) opportunities, challenges, and risks of AI, (2) conditions for the responsible use of AI, (3) potential future scenarios, and (4) prioritized actions. The first Delphi wave followed an open and exploratory approach, while the second wave sought to prioritize and rank key findings from the previous round. The survey instruments included open-ended and closed-ended questions that were analyzed through statistical analysis, inductive and deductive coding. Results Results showed that AI holds various opportunities for the RCCE-IM of public health communicators. Challenges and risks, such as algorithmic bias or concerns about inclusivity, equally affect RCCE-IM efforts. Experts identified seven principles for the responsible use of AI (e.g., fair, human-centered) and predicted optimistic (e.g., optimized workflows) and pessimistic (e.g., flood of (mis)information) scenarios for the future impact of AI. Prioritized actions range from regulation, resource allocation, and feedback to capacity building, public trust, and education. Conclusions To responsibly navigate the opportunities, challenges, and risks of AI for public health emergency communication, clear guiding principles, ongoing critical evaluation, and multi-sectoral collaboration are needed. Key messages • AI is likely to transform public health emergency communication. • A set of agreed-upon principles will support public health authorities in the responsible management of the opportunities, challenges, and risks of AI.