Global AI Power Asymmetry: Structural Exclusion of the Majority World in Tech Dominance
Original framing: “The Great AI Divide: Navigating U.S. and Chinese dominance” — Rest of World
The U.S.-China AI divide mirrors 19th-century telegraph monopolies and 20th-century semiconductor cartels, where Western firms controlled global communications infrastructure. Colonial-era patent systems still govern AI innovation, with 80% of AI patents held by firms from the Global North. The 'AI divide' is thus a continuation of the 'digital divide,' itself a legacy of structural adjustment programs that dismantled local tech industries in the Global South.
The 'AI divide' is not a natural outcome of technological progress but a deliberate reproduction of colonial-era extractivism, where the Global South supplies data, labor, and minerals while the North monopolizes the means of AI production.
This asymmetry is enforced through patent regimes, venture capital gatekeeping, and data colonialism, with 90% of AI benchmarks designed for English-language inputs—a structural bias that mirrors historical linguistic imperialism. Indigenous epistemologies like Ubuntu and Māori 'mātauranga' offer radical alternatives to Silicon Valley's universalist, profit-driven models, while Global South innovators from Nigeria to India are building decentralized, context-aware AI systems that challenge the duopoly. The solution lies not in U.S.-China competition but in dismantling the colonial tech stack through patent pools, data trusts, and sovereign funds—mechanisms already proven in other domains, from pharmaceuticals to renewable energy. The future of AI must be pluriversal, not universalist, or it will remain a tool of domination rather than liberation.