Practising Algorithmic Authenticity in East African Public Relations: A Framework Grounded in Kenyan Cases

Abstract

Generative artificial intelligence has moved rapidly from novelty to everyday instrument in East African public relations, strategic communication, and institutional visual production. Adoption is largely ungoverned, and the tools carry the aesthetic, linguistic, and cultural defaults of the Global North data on which they are trained — an issue of algorithmic justice with direct consequences for media integrity in the region. This paper reconceptualises authenticity, long treated as the currency of public trust in PR, as an ongoing practice rather than a fixed property a message possesses or lacks: a set of deliberate ethical decisions practitioners make each time they engage an AI tool. Drawing on sixteen years of PR practice and teaching in Kenya, the paper proposes a four-dimensional framework — disclosure, cultural-linguistic fidelity, human accountability, and proportionality — and illustrates it through Kenya as an information-rich empirical entry point into a broader East African and Global South dynamic. Two documented Kenyan cases (Posta Kenya's AI-generated customer-service image and the Kenya Urban Roads Authority's culturally implausible road imagery) anchor the analysis, read through visual social semiotics and a decolonial lens. A corroborating case — the US State Department's AI-generated map at the AIDS 2026 conference, which misrepresented Uganda and other African countries — extends the framework beyond the region's own institutions. Faith-based communication is noted as a setting where the framework's questions sharpen further, contributing a locally grounded, teachable framework for practitioners, educators, and institutions across East Africa.