Without intentional strategies to strengthen energy infrastructure, the Global South will remain largely a consumer of AI technologies.

Five years after the launch of the UN Roadmap for Digital Cooperation, the global conversation on artificial intelligence is beginning to shift beyond the advanced economies. Reflecting on India’s plan to host the AI Summit in 2026 —  the first such summit to be held in a developing country —  the UN Secretary-General’s Tech Envoy highlighted how rapidly the geography of AI governance is evolving.

This moment underscores a broader shift: while artificial intelligence is reshaping global economic power, the benefits are accruing unevenly. So far, the rapid rise of AI has positioned the Global North and China to capture outsized economic and technological gains. Meanwhile, the Global South — home to over 70 percent of the world’s population — risks being left behind. The central constraint is not ambition, data, or talent alone, but something far more basic: access to affordable, reliable, and clean electricity. Without intentional strategies to strengthen energy infrastructure, the Global South will remain largely a consumer of AI technologies, deepening structural inequalities and missing a critical opportunity to harness AI for growth and development.

A Data-Rich, Infrastructure-Poor Global South

Paradoxically, the Global South is becoming increasingly data-rich even as it remains infrastructure-poor. The rapid expansion of mobile connectivity has turned many developing economies into major generators of digital data. New subsea cable projects — linking Africa to Europe and Asia to the United States — have dramatically increased bandwidth, reduced latency, and improved access for remote geographies. Sub-Saharan Africa alone is projected to add 167 million new mobile subscribers by 2030. 

Yet this explosion of data has not translated into local value creation. With limited domestic data-processing infrastructure, much of this data is exported to foreign data centers for storage and analysis. Accounting for roughly 50 percent of the world’s internet users (excluding China) but only about 10 percent of global data center capacity, the Global South has become a net exporter of data. As AI applications scale, this imbalance risks becoming entrenched — producing models that neither reflect the Global South’s realities nor serve local priorities.

At the heart of this imbalance lies a physical constraint: power. Data centers — the backbone of AI development — require uninterrupted, high-quality electricity. Yet many Global South countries operate fragile grids prone to outages and load shedding. Introducing energy-intensive AI infrastructure in such contexts risks intensifying competition with manufacturing and essential services for the limited grid capacity. The solution, however, is not to replicate legacy power systems built for centralized, predictable demand. Instead, investing in flexible, digitalized, and renewable-ready grids can create a decisive competitive advantage.

Smart grid and off-grid pilots in Kenya and across East Africa have demonstrated how renewables can be integrated at scale while improving access and reliability. Improved connectivity now also enables data centers to be sited away from population centers and closer to renewable resources, reducing pressure on urban grids. Data-center-specific strategies — such as on-site renewable generation, microgrids, and battery storage — offer pathways to leapfrog outdated infrastructure. India’s solar parks, Brazil’s large-scale renewables, and South Africa’s wind corridors could all anchor emerging AI ecosystems.

The Key Ingredients to Data Center Domination: Economics, Infrastructure, and Skilled Labor

The economics increasingly favor this approach. The combination of plummeting renewable energy costs and flexible grid architectures makes it possible to power AI infrastructure while simultaneously strengthening energy security. By contrast, reliance on imported coal or natural gas exposes AI ambitions to volatile global fuel markets and geopolitical risk. Domestic and on-site renewables can insulate both industrial and residential consumers from price shocks, reducing social and political risk. Brazil’s northeast, for example, is scaling renewable capacity alongside energy-intensive industries at globally competitive costs. In India, solar projects in Rajasthan — paired with battery storage — are being designed to supply industrial clusters with reliable power, laying the foundation for AI-ready infrastructure.

Energy infrastructure alone, however, is not sufficient without also having the regulatory systems that enable rapid deployment. Even the most advanced grids and cheapest power will fail to attract AI investment if permitting and interconnection processes are slow or opaque. This challenge is increasingly visible in Global North data center hubs, where permitting delays and grid connection queues have become binding constraints. By contrast, several Global South countries are experimenting with more agile regulatory approaches. Vietnam’s direct power purchase agreements (DPPAs), for instance, allow technology firms to source renewable electricity directly from private developers, reducing risk and accelerating timelines. 

Following recent permitting reforms, new opportunities are also emerging for co-investment between global technology firms and local energy developers. Google’s record investment in Visakhapatnam, India — bundling data centers, subsea cables, renewable capacity, and grid upgrades — illustrates how integrated regulatory frameworks can unlock economies of scale. Countries that offer clear, predictable, and interoperable rules will increasingly act as magnets for AI investment.

Finally, neither energy systems nor AI infrastructure is self-sustaining without skilled human capital. Building and operating future-ready grids and data centers requires technicians, engineers, and system operators with new skill sets. In the near term, retraining existing energy sector workers to deploy and maintain renewable assets is essential. Over the longer term, universities and technical institutes must produce professionals capable of managing advanced grid systems and AI-enabled operations. At the same time, automation will continue to reshape these roles, raising difficult questions about workforce displacement. Continuous upskilling and reskilling will be critical to ensure that local communities benefit from the industries they help create, rather than being sidelined by them.

The Global South now stands at a narrow but consequential inflection point. Without proactive action to modernize grids, scale renewables, streamline regulation, and invest in human capital, the AI boom will bypass much of the developing world and exacerbate global inequality. Acting quickly will involve tradeoffs and disruption. But the potential gains — energy independence, economic competitiveness, and accelerated development — are transformative. If the Global South wants a seat at the AI table, it must start with the fundamentals: turning the lights on — and keeping them on.

About the Author: Medha Prasanna

Medha Prasanna is program coordinator and junior fellow at ORF America, where she contributes to research, writing, and programming on energy, climate, and sustainability. She tracks energy and climate developments between the United States and India, two of the world’s largest economies and energy consumers. She received an M.A in International Affairs from the Elliott School of International Affairs at the George Washington University and was a recipient of their Global Initiative Fellowship.

Image: Shutterstock/sdecoret

The post Grids Will Decide the Global South’s AI Future appeared first on The National Interest.