WHO and artificial intelligence: contesting global health futures through foresight

TL;DR

The paper analyzes how the World Health Organization discusses artificial intelligence in global health, focusing on how it frames AI's potential and influence.

Contribution

It introduces a novel analysis of WHO's AI discourse through strong and weak AI narratives, revealing strategic framing and governance priorities.

Findings

WHO's foresight exercises emphasize weak AI narratives focusing on technical limitations and ethical concerns.

The WHO strategically isolates AI from other technologies while blurring boundaries to legitimize its policy authority.

Expert participation in WHO's foresight exercises helps validate specific AI trajectories aligned with existing priorities.

Abstract

The article examines the World Health Organization’s (WHO) discourse on artificial intelligence (AI) in their foresight exercises, doing so by drawing on the analytical framework of strong and weak AI narratives. The analysis finds that strong AI narratives (those which depict AI as human-like or even super-intelligent, emphasising existential risks and transformative power) are rarely found. In contrast, the exercises produce a broad range of weak AI narratives (those that emphasise the technical limitations, ethical concerns, and practical governance of specific AI applications in healthcare). The findings reveal how certain AI technologies are foregrounded by WHO, how these are framed as in isolation from other emerging technologies, how this isolation is strategically blurred, and the role of expert participation in legitimising WHO’s policy on AI. Situated within WHO’s broader policy discourse on AI, the paper draws out how the foresight exercises strategically construct and validate particular trajectories aligned with WHO’s existing priorities. Through selective narrative framing, expert input, and methodological design, WHO reinforces its epistemic authority by guiding the discourse on AI in global health toward context-sensitive and manageable use cases of the technology. Ultimately, these foresight exercises serve as a site of contestation, where competing visions of AI in global health are negotiated, and WHO’s influence over future governance in the area is actively shaped.

Figures