Responsible AI-driven public health communication in conflict zones: A case study of Palestine

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Public health communication in conflict zones faces unique and formidable challenges, including the collapse of healthcare infrastructure, the proliferation of misinformation, and deep-seated sociopolitical distrust. This paper critically examines the potential and limitations of artificial intelligence (AI), particularly large language models (LLMs), in strengthening public health communication under such conditions, with Palestine as a case study. While AI-driven systems offer new possibilities for multilingual, real-time, and culturally tailored messaging, they also introduce serious risks, including hallucination, bias reproduction, and the exacerbation of infrastructural inequalities. To address these complexities, the paper proposes the development of a World Health Organization-supervised health LLM trained on verified datasets and governed by strict ethical protocols. It also advocates for the inclusion of AI literacy as a foundational competency for humanitarian health workers. Through a detailed exploration of ethical, infrastructural, and sociological dimensions, the paper argues that responsible AI deployment must prioritize trust-building, cultural sensitivity, and equity to enhance health resilience in conflict-affected societies.