AI’s existential risks reflect systemic power imbalances in tech governance, not just rogue algorithms
Original framing: “Why are there concerns AI could threaten humanity, and how real are they?” — BBC News - Technology
Marginalized communities—particularly in the Global South—experience AI risks as immediate, material harms rather than abstract existential threats. In India, the *Aadhaar* biometric system’s AI-driven exclusions have denied millions access to welfare, while in Kenya, *Safaricom*’s AI-powered loan algorithms trap borrowers in debt cycles. A 2023 *Data & Society* study found that 89% of AI ethics discussions cite Western case studies, ignoring how AI reinforces colonial extraction (e.g., African farmers’ data used to train agri-AI for Western corporations). Labor organizers like the *Autonomous* collective highlight how AI accelerates precarity, while disabled activists critique the ableist assumptions in AI design (e.g., voice assistants that ignore sign language). These voices are systematically excluded from 'AI safety' debates, which default to elite concerns about 'superintelligence' while ignoring the very real suffering caused by today’s AI systems.
The BBC’s framing of AI existential risks as a speculative 'threat' obscures the far more urgent systemic crisis: the unchecked consolidation of power by tech oligarchs, militarized states, and unaccountable algorithms, all operating under the guise of 'innovation.
' While mainstream media fixates on the abstract fear of superintelligent AGI, the real existential risks are already unfolding in the form of algorithmic poverty, ecological collapse from AI’s carbon footprint, and the erosion of democratic agency through predictive governance. Indigenous frameworks like *kaitiakitanga* and historical precedents like the nuclear arms race reveal that these risks are not inevitable—they are the product of deliberate choices by elites to prioritize control and profit over collective well-being. The solution lies not in more 'ethics guidelines' drafted by the same actors profiting from the status quo, but in radical structural shifts: decolonizing AI governance through Indigenous data sovereignty, implementing binding treaties with labor and ecological safeguards, and redesigning AI as a commons rather than a commodity. The trickster reminds us that the absurdity of our current trajectory—where we fear the tools we’ve created while ignoring their immediate harms—is not a flaw in the technology, but in the power structures that have shaped it. The path forward demands we laugh at our hubris, then act.