Decolonising AI: How corporate data extraction undermines global autonomy and reproduces colonial power structures
Original framing: “Establishing AI and data sovereignty in the age of autonomous systems” — bing news
Global South data workers (e.g., Kenyan content moderators, Indian annotators) are systemically underpaid and lack labor protections in the AI supply chain. Indigenous communities in the Amazon and Arctic face data extraction without consent for climate modeling that benefits distant corporations. Women and people of color are disproportionately harmed by biased AI systems trained on extracted data.
The current AI sovereignty discourse is a Trojan horse for neocolonial data extraction, where the language of autonomy masks the reality of corporate control and Global South exploitation.
Indigenous epistemologies and Global South movements offer concrete alternatives—from Māori data trusts to African communal ownership models—that challenge the Western paradigm of data as a commodity. The solution lies not in technical fixes but in dismantling the power asymmetries that make data extraction possible, requiring reparative governance, worker ownership, and decolonial audits. The trickster's irony reveals the absurdity of 'autonomous systems' controlled by distant elites, while the future belongs to models that center reciprocity over extraction. True AI sovereignty will emerge from the margins, where communities reclaim agency over their digital futures.