Beyond ‘data colonialism’ — The role of Global South in an AI world
For AI to serve the world, the early threads of compute and capital in Global North must be woven with the texture of humanity and language of the Global South.
The world has never been evenly arranged. Borders, ideas, capital, opportunities have never ever been equally distributed. The language of global development has always attempted to soften this reality with polite abstractions such as “developing countries” or “emerging economies”, yet the deeper historical geography of inequality is best captured in one term: ‘Global South’.
Global South captures an evolving community of countries across Asia, Africa, Latin America, the Middle East, and the Pacific that share a common experience of historical marginalisation under colonial rule, exclusion from the centres of power and uneven participation in the global economy. The discussion becomes especially significant in the age of Artificial Intelligence (AI) — a domain that threatens to reproduce, and potentially worsen, the very asymmetries that the Global South has long struggled with and faced barriers.
Data asymmetry is the first barrier. AI systems are trained on enormous datasets, but much of this training data originates from global digital platforms dominated by northern companies. The lived realities of the Global South, like its languages, dialects, cultural practices, informal economies, agricultural rhythms, local wisdoms are underrepresented. This leads to what researchers call data colonialism: a condition where the Global South contributes raw data but receives little agency in shaping the systems built from it.
Compute asymmetry is the second barrier. Training large AI models requires specialised hardware and high-density data centres that are expensive to procure, energy-intensive to run, and logistically complex to maintain. Most of the Global South does not have sovereign compute infrastructure at scale.
Research asymmetry is the third. AI research output is overwhelmingly concentrated in institutions in the North. Scientists and engineers from Global South contribute talent, but often migrate out due to better funding and professional opportunities abroad.
And finally, there is governance asymmetry. AI regulation frameworks on safety, privacy, fairness and accountability are being written predominantly in the North. Without strong representation from Southern governments, the resulting norms risk embedding the priorities, anxieties and political structures of wealthy nations while overlooking the realities of the world’s majority.
Together, these asymmetries create a troubling possibility that the age of AI may reproduce the same patterns of extraction and dependency that defined the colonial and post-colonial eras. Yet this story is not inevitable. The Global South is a geography of scale, diversity, ingenuity, and resilience. And it contains capacities that the world’s AI future cannot do without.
- First, the Global South holds the next billion users and a demographic and economic force that will shape digital ecosystems for decades to come.
- Second, the Global South is technologically ambitious. Countries like India, Brazil, Indonesia, South Africa and Vietnam are building digital public infrastructure at a pace unimaginable two decades ago. Unified payment systems, digital identity frameworks, e-governance stacks, community health networks are fertile ecosystems where AI can be responsibly embedded at population scale.
- Third, the Global South has a demographic that's young, creative, entrepreneurial and digitally native. Innovation often emerges where constraints are high and resources limited. The South’s frugal engineering, improvisational genius and problem-first mindset offer alternative pathways to AI development which are grounded in societal need and not merely commercial optimisation.
To leverage the potential, the South must invest in four strategic pillars.
1. Data Sovereignty: For decades, data from the Global South has flowed outward and captured by global platforms, monetised elsewhere and rarely reinvested into local capabilities.
This cycle must be broken. It means ensuring that national datasets remain accessible for local innovation and designing consent, privacy and governance frameworks that reflect local cultural norms.
India’s Digital Public Infrastructure (DPI), Brazil’s open banking frameworks and Africa’s cross-border digital identity experiments show that the Global South can create robust, citizen-first data ecosystems that outperform purely private models.
2. Compute Equity: At present, the world’s high-end computing resources like GPU clusters, training supercomputers and hyperscale data centres are in the US, China, and a small number of wealthy economies. For the Global South to participate meaningfully in AI, it must overcome this bottleneck through regional compute clouds, funded cooperatively by governments.
The computing clusters, whether private or public must be made accessible to startups, universities and social-sector organisations. South-South technology alliances should be conceived.
3. Distributed Research Capacity: AI research today is dominated by a small cluster of institutions in North America, Europe and China. The Global South must build its own distributed network of AI research centres that specialise in context-rich problems. This requires creating AI excellence centers focused on agriculture, climate, public health, education and language technologies.
The Global South needs to reverse brain drain through fellowships, diaspora return programs and cross-border collaborations
4. Governance Coalitions: For AI governance to be legitimate, the Global South must shape it. This requires coalitions within the UN, OECD and UNESCO that articulate Global South priorities.
Conclusion
The Global North may hold the early threads of compute and capital, but the Global South holds the textures of humanity in the languages, the communities and the lived realities that give meaning to life. For AI to serve the world, the world must weave it together. And in that weaving, the Global South is not a latecomer. It is an essential co-author of the future.
The author, Sudhir Tiku, is a technical philosopher and AI Influencer based out of Singapore. He has over two decades of global work experience in automation and analytics. He is a regular at the TEDx circles in Asia Pacific and is a passionate traveler and blogger.
