The Infrastructure Paradox of AI Development in LMICs
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The regions that stand to benefit most from artificial intelligence are often the least equipped to develop it locally, a tension we term the "Infrastructure Paradox." This exploratory, cross-sectional study surveys 91 AI practitioners across 27 low- and middle-income countries (LMICs) to assess the availability, affordability, and reliability of four foundational AI infrastructure pillars: electricity, internet connectivity, computing hardware, and training data. Rather than a uniform deficit, the results reveal an income-stratified "Infrastructure Gradient" characterised by three non-linear dynamics: a Reliability Trap, wherein lower-middle-income countries report higher electricity service failure rates than low-income countries; a Data Quality Divide, where data accessibility remains constrained across income groups but quality improves markedly with income; and a Hardware Inversion, wherein AI interventions concentrate where hardware is most expensive, suggesting external financing rather than sustainable local markets drives current AI activity. Policy presence correlates with intermediate infrastructure capacity, implying governance frameworks emerge where capability and urgency coincide. These findings reframe infrastructure as a governance and equity determinant shaping where, how, and for whom AI develops, and suggest that technological leapfrogging in AI is conditional rather than categorical.
Fields of Study & Subjects
Ethics and Social Impacts of AI
ICT Impact and Policies
ICT in Developing Communities
Publication details
- Journal / source
- Research Square
- Publication date
- 31 March 2026
- Language
- EN
- Document type
- Preprint