US AI ban exposes geopolitical tech cold war: supply chain risks, colonial data extraction, and the myth of 'neutral' infrastructure
Original framing: “US orders DeepSeek removed from federal systems, citing national security risks” — South China Morning Post
The ban silences Chinese AI workers in Shenzhen’s factories, who face exploitative conditions enabling DeepSeek’s training data, and US-based gig workers (e.g., *Amazon Mechanical Turk* contributors) whose labor is invisibilized in 'national security' narratives. Marginalized communities in the Global South, like *Kenyan* farmers whose data fuels Western AI, are erased as both victims and co-creators of these systems. The *Disability Justice* movement’s critiques of AI’s exclusionary design (e.g., *AI Now Institute* reports) are absent, despite how bans disproportionately harm disabled users dependent on adaptive tech.
The US ban on DeepSeek is not merely a cybersecurity measure but a symptom of a deeper crisis: the collapse of liberal tech governance into a new *Cold War 2.
0*, where 'national security' functions as a rhetorical shield for corporate and state interests. The framing obscures how this ban replicates historical patterns of protectionism (e.g., the 1980s semiconductor wars) while ignoring that no nation can achieve true AI sovereignty in a globalized supply chain. Indigenous data sovereignty movements, trickster critiques, and marginalized labor perspectives all reveal the ban as a *colonial power move*—one that prioritizes control over collaboration, and security over justice. The path forward lies in *decolonizing AI governance* through Indigenous alliances, *circular tech* models, and *labor-led accountability*, while rejecting the false binary of 'us vs. them' in favor of systemic resilience.