AI, Media, and the Protection of Marginalised Voices in Nigeria: A Conceptual Analysis

Abstract

This paper offers a conceptual and normative analysis of how artificial intelligence is reshaping the production, circulation, moderation, and monetisation of media content in Nigeria. It offers speed, reach, translation, automation, and new creative possibilities, yet it can also weaken the truth-seeking role of the media by normalising synthetic content, thinning editorial judgment, and rewarding virality over verification. The paper argues that the central question is not whether AI can make communication more efficient, but whether AI-driven media systems in Nigeria will deepen or reduce structural and epistemic marginalisation. Structural marginalisation concerns unequal access to visibility, credibility, and amplification. Epistemic marginalisation concerns whose forms of speaking, reasoning, and knowing are recognised as meaningful in a multilingual society where local-language, oral, code-switched, and proverb-rich expression is often sidelined by English-dominant digital systems. Drawing on journalism studies, platform governance, algorithmic bias research, African communitarian ethics, and selected Western ethical traditions, the paper develops an integrated ethical framework through analytical synthesis rather than empirical fieldwork. It argues that welfare, dignity, reciprocity, virtue, solidarity, and participation must be considered together. On that basis, it proposes six practical commitments for AI in media in Nigeria: transparency, verification, inclusion, accountability, participation, and media literacy. The conclusion warns that, unless governance improves, Nigeria may move toward an increasingly stratified media ecology in which machine-legible, English-dominant content gains disproportionate institutional visibility while local-language and oral forms remain underheard. At the same time, the paper insists that translation and accessibility tools can help only when they are embedded in a broader politics of epistemic recognition rather than treated as technical substitutes for it, and it identifies priorities for future empirical research on local-language moderation, code-switching, oral communication, and synthetic media in fragile settings.