Artificial Intelligence for Public Health Surveillance: Promises, Pitfalls, and an Ethics-Equity Roadmap for Deployment
Disease Prevention and Epidemiology · 12.01.2026 · 1 (1) · 22-39
DOI: 10.64229/d36s6961 ↗Kurz gefasst
Fachartikel in der Zeitschrift Disease Prevention and Epidemiology, online erschienen am 12.01.2026. Verfasst von Gavin Chibundu Ikechukwu und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Themen: Künstliche Intelligenz in der Pflege, Ethik in der Pflege
Abstract (Originalsprache)
Artificial intelligence (AI)__such as machine learning (ML), natural language processing (NLP) and large multimodal models-is transforming the field of public-health surveillance, facilitating earlier outbreak detection, the combination of heterogeneous streams of data, and high-scale signal triage. All these capabilities have potential to enhance timeliness, sensitivity and geographic coverage of the surveillance systems, especially when used together with nontraditional sources of information like social media, news media, electronic health records, mobility data and environmental (e.g. wastewater) indicators. Though, actual implementations reveal significant traps: biased training data may enhance health inequities, bad data management and inadequate models may harm trust and responsibility, and incomplete validation may lead to false alarms or missed events. Moreover, the use of surveillance also presents more ethical issues, including privacy and consent to surveillance and the likelihood of surveillance increasing social evils. The review summarizes the existing evidence and practice in the technical, ethical and governance aspects of AI in surveillance of public-health. We list recent achievements and opportunities, methodological and operational issues, and trace the most significant risks. Based on global best practices and new reporting recommendations, we come up with a realistic roadmap towards ethical, equitable, and successful AI surveillance implementation. Among the recommendations, there are: strict pre-deployment validation on representative datasets; model documentation and performance reporting (where applicable) based on TRIPOD+AI/CONSORT-AI, continuous concept drift monitoring and measures of fairness, robust data governance and privacy-preserving architecture, stakeholder involvement, including affected communities, and building capacity in low- and middle-income countries (LMICs). We conclude by outlining priority research and policy measures that are required to ensure the translation of the technical promise of AI into benefits to the public-health with minimal harms.
Zitieren
APA
Ikechukwu, G. C., Egba, S. I., Paul, P. C., & Umeh, V. A. (2026). Artificial Intelligence for Public Health Surveillance: Promises, Pitfalls, and an Ethics-Equity Roadmap for Deployment. Disease Prevention and Epidemiology, 1(1), 22-39. https://doi.org/10.64229/d36s6961
Vancouver
Ikechukwu GC, Egba SI, Paul PC, Umeh VA. Artificial Intelligence for Public Health Surveillance: Promises, Pitfalls, and an Ethics-Equity Roadmap for Deployment. Disease Prevention and Epidemiology. 2026;1(1):22-39. doi:10.64229/d36s6961
BibTeX
@article{Ikechukwu2026,
author = {Ikechukwu, Gavin Chibundu and Egba, Simeon Ikechukwu and Paul, Promise Chibuike and Umeh, Virginus Agozie},
title = {Artificial Intelligence for Public Health Surveillance: Promises, Pitfalls, and an Ethics-Equity Roadmap for Deployment},
journal = {Disease Prevention and Epidemiology},
year = {2026},
volume = {1},
number = {1},
pages = {22--39},
doi = {10.64229/d36s6961}
}
EUVIMED