Will AI really become universal? The colonial roots of machine learning

Written by Bhavani Shankar Nayak*

Artificial intelligence is no longer a futuristic fantasy. It is rapidly becoming the invisible architect of our daily lives, influencing what we eat, who we make friends with, how we heal, and where we work. In the coming years, AI will not only facilitate human choices; That will make them more and more determined. This inescapability presents an obvious paradox. AI has the potential to deepen democracy, decentralize power, and decolonize knowledge, but it also threatens to entrench the very inequalities it has the potential to eliminate.

The danger lies not in the technology itself, but in the data that powers it. Most AI systems and their large-scale language models (LLMs) are trained on datasets drawn overwhelmingly from Europe. Please consider this. Approximately 90 percent of the World Digital Library’s archival records come from Europe, and only 10 percent from the rest of the world. When AI learns from such biased sources, Eurocentric bias is a feature, not a bug. These systems do not simply reflect a Western worldview. They marginalize the indigenous knowledges of Africa, Asia, and Latin America as mere folklore and ethnography, while elevating them to the status of universal scientific truth.

This epistemic violence is no accident. The architecture of AI itself, its reliance on binary “input” and “output” logic, reflects Cartesian dualism, a framework that has historically dismissed non-European knowledge traditions as inferior. By building this duality into its core, AI inadvertently reproduces colonial hierarchies and suppresses the promise of its own liberation. The goals of democratization, decolonization, and even decarbonization are undermined when technologies that are supposed to benefit all humanity only benefit a narrow, profit-seeking elite.

The impact is already visible. Economically, AI facilitates the platform markets of digital capitalism, reducing social relationships to transactional data. Politically, it concentrates power in the hands of a small number of platform capitalists, enabling a quiet authoritarianism to dominate everyday life. Culturally, it promotes mass consumerism and standardization, flattening food, entertainment, and even dreams into homogeneous products. Capitalism promised choice. AI enables conformity.

Worse, AI reduces humans to data points. Our wants, needs, and desires are stripped of their context and treated as variables in a causal equation that ignores the complex intergenerational conditions that shape true happiness and peace. AI has revolutionized productivity and even warfare, but it has tragically proven incapable of understanding the richness of human existence.

The solution is not to abandon AI, but to decolonize it. That means fundamentally diversifying the datasets, epistemologies, and governance structures that underpin these systems. AI needs to be collectively owned, operated, and guided by authentic multiple knowledge traditions around the world. Only then can its democratic potential be realized, not as an instrument of domination, but as a catalyst for equality, justice, and true human flourishing.

—

*UK based academic

Source link