AI Is Recreating a Colonial World Order

The future of AI is increasingly described as a race between Washington and Beijing. Export controls, chip restrictions, and competing infrastructure projects now dominate the headlines. But the real contest in AI is not who wins the U.S.–China race. It is who gets to shape the systems everyone else must live with.

The deeper transformation already underway is that the AI economy is entrenching a global hierarchy. Silicon Valley, Europe, and China are positioned as sites of invention, while the Global South is cast as a site of labor and deployment – a colonial pattern now reproduced through data, compute, and capital.

Across Africa and Asia, millions of workers already help train AI models as coders, content moderators and data labelers, for companies like OpenAI and Samsung – often for as little as $2 per hour. Reports of poor labor practices are rife. San Francisco-based subcontractor Sama and its client Meta are facing a class action lawsuit for exploiting content moderators in Kenya. More than 140 workers claim that their role forced them to constantly view atrocities – including violence and abuse – that left them struggling with PTSD, depression, or anxiety.

Billions continue to flow into experimental Silicon Valley models. Yet, when builders from the Global South attempt to create systems of their own, they encounter a very different set of conditions. African startups receive less than 1–2% of global venture capital despite representing nearly a fifth of the world’s population, while operating under infrastructure constraints that would be unthinkable in established markets. Capital flows reflect more than risk—they encode assumptions about where credible innovation is expected to come from.

Organizations like Neem Capital—an AI-native investment firm connecting global capital with high-impact opportunities in emerging markets—argue that this bias is structural. As its co-founder Mo Marikar explains, “Founders outside established innovation hubs are often expected to prove legitimacy before they are even evaluated on merit.”

This dynamic becomes most visible when companies from beyond traditional tech centers attract global attention. It shaped the reception of InstaDeep, a Tunisia-founded AI firm later acquired by BioNTech. Despite developing advanced reinforcement learning systems applied to logistics, genomics, and pandemic response, its technical contributions were often secondary in coverage to the fact of its acquisition.

What transformed InstaDeep from a regional player into a globally credible AI company was not a sudden shift in its technology, but its incorporation into a European firm. For companies emerging from Africa, technical achievement alone is rarely sufficient to establish legitimacy.

These credibility gaps are reinforced by material ones. Governments across Africa have already spent more than $2 billion embedding Chinese AI-enabled surveillance systems into public infrastructure, while U.S. firms entrench dominance through cloud platforms and proprietary foundation models. Local economies increasingly run on infrastructure they do not own. Public decisions rely on systems they cannot fully audit. Data generated locally is extracted, processed elsewhere and monetized abroad.

This is not a gap the market will correct on its own. Systems built on narrow assumptions about who gets to invent will produce narrower outcomes. They will embed bias into technical standards, limit the diversity of solutions available to emerging markets and increase dependence on infrastructure controlled elsewhere. In an interconnected global economy, those distortions do not remain regional problems for long.

Some argue that this concentration of capability is simply the natural outcome of scale. Advanced AI systems require enormous compute resources, specialized talent pools and sustained investment. But that explanation mistakes a political settlement for an economic inevitability. Today’s hierarchy is being built through decisions about where capital flows, whose claims are trusted and which ecosystems are allowed to mature.

Addressing this imbalance requires more than expanding access—it requires rethinking how credibility and capital are assigned in the first place. That means backing institutions and investors willing to fund AI ecosystems in emerging markets on their own terms, rather than as extensions of existing hubs. However, scaling that model will require a broader shift: from viewing the Global South as a source of labor to recognizing it as a site of innovation.

Ultimately, AI systems are already reshaping how credit is allocated, how borders are managed, how health systems triage patients and how governments monitor citizens. If the authority to design those systems remains concentrated in a handful of regions, then their assumptions about risk, identity, language, and value will quietly become global defaults.

A global AI economy that excludes much of the world is not just unjust—it is unstable.

Therefore, the defining question of the AI era is not whether Washington or Beijing leads the next generation of models. It is whether the infrastructure shaping the future of work, governance, and opportunity will be built by a narrow set of actors—or by a genuinely global community of innovators.

Because if credibility continues to follow old geographic lines, AI will not just reproduce inequality—it will formalize it.

* Aaliyah Vayez is an international relations analyst specializing in geopolitical risk and global power competition. Her writing has appeared in BBC Africa, the Guardian, and The Conversation, among others. She advises multinational firms on emerging market risk and holds an MSc in International Relations from the London School of Economics.

Source: https://znetwork.org/znetarticle/ai-is-recreating-a-colonial-world-order/