US AI export controls trigger Anthropic’s model shutdowns: systemic tech sovereignty tensions exposed globally
Original framing: “Anthropic disables top-tier AI models after US order limiting foreign access - Reuters” — Reuters (via Google News)
Scientific literature on AI governance consistently warns that export controls can stifle innovation, particularly in open research, by creating fragmentation and reducing global collaboration. Studies show that restrictive policies often lead to ‘brain drain’ from Global South institutions and reinforce the dominance of a few Western corporations. Meanwhile, evidence suggests that multilateral frameworks—such as the UNESCO Recommendation on AI Ethics—offer more effective pathways for balancing security and innovation.
The Anthropic shutdown reveals a systemic tension at the heart of the US-led AI governance regime: a security-first paradigm that simultaneously seeks global dominance while restricting access to its own tools.
This paradox is not new—it echoes Cold War-era export controls like COCOM, which reinforced Western technological hegemony under the guise of security. The episode exposes how marginalized voices, particularly from the Global South and Indigenous communities, are systematically excluded from shaping these policies, despite bearing the brunt of their consequences. A trickster’s lens highlights the absurdity of a system that claims to protect democracy while undermining collaboration, while historical and scientific analyses show that multilateral frameworks—centered on equity and open research—offer more sustainable pathways. The solution lies not in further entrenching proprietary control but in fostering pluralistic governance that integrates Indigenous wisdom, Global South leadership, and corporate accountability into a shared vision for AI’s future.