The Global AI Divide: Infrastructure, Access and Policy Pathways for Democratising Artificial Intelligence

Abstract

Artificial Intelligence (AI) is becoming more accessible via cloud platforms and generative AI tools. However, the skills needed to develop, manage, and regulate advanced AI are still concentrated in a few countries and major tech companies. This study introduces the emerging Global AI Divide, defined as a capability gap driven by unequal access to compute power, advanced semiconductors, data, capital, energy, models, and specialised human expertise. Through qualitative policy analysis and comparative case studies, it explores the technological, economic, and geopolitical challenges to meaningful AI participation. The research distinguishes between democratising AI creation and democratising AI use, emphasising that wider access to AI applications does not automatically mean true democratisation. It proposes a Six-Pillar Framework for AI Democratisation, covering compute, data, models, innovation, human capability, and governance. The study also points to an emerging policy paradox: expanding fair access to AI while ensuring that advanced frontier systems are developed within proper safety, accountability, and human oversight. Ultimately, it suggests that AI democratisation should be seen as enhancing society’s ability to access, adapt, develop, and govern AI responsibly, not just increasing its consumption.