Language inequality in artificial intelligence: Nigerian languages, AI underrepresentation, and digital inclusion

Abstract

Artificial intelligence increasingly shapes access to information, Education, public services, civic communication, and economic opportunity. However, AI systems do not support all languages equally. This paper examines the underrepresentation of Nigerian languages in artificial intelligence and its implications for digital inclusion. It has three objectives: to examine how selected Nigerian languages and language varieties are represented in AI and NLP systems; to assess how selected public-facing AI tools respond to English, Nigerian-language, and mixed-language examples; and to discuss the implications of language inequality for digital inclusion, public services, Education, civic participation, and economic opportunity. Focusing on Yorùbá, Igbo, Hausa, and Nigerian Pidgin as illustrative cases, the paper employs conceptual analysis, secondary evidence from African natural language processing research, and a small exploratory assessment of AI tools, including ChatGPT, Google Gemini, and Microsoft 365 Copilot Chat. The findings suggest that language inequality in AI operates across data availability, model representation, and user-facing performance. Although selected Nigerian languages appear in emerging African NLP resources, their presence does not guarantee robust support for public-facing use. The AI-tool assessment showed that English responses were more stable and complete. In contrast, Nigerian-language and mixed-language outputs exhibited greater variation in meaning preservation, register, cultural context, information completeness, and public usefulness. The paper argues that Nigerian-language underrepresentation is both a technical problem and a governance issue shaped by low-resource conditions, limited datasets, market incentives, infrastructure constraints, and postcolonial language politics. It concludes that linguistically inclusive AI in Nigeria requires stronger datasets, community-led evaluation, policy support, sustainable funding, and AI governance frameworks that treat Nigerian languages as part of digital infrastructure.