The release of DeepSeek’s R1 artificial intelligence model and Moonshot AI’s Kimi K2 has sent shockwaves through policy circles in the United States. China’s bid to become the world’s leading distributor of AI capabilities is facing several hurdles, but the US must adapt quickly to maintain its global influence.
The Chinese company DeepSeek released its R1 AI model, which has gained significant traction worldwide, despite U.S. export controls on advanced semiconductors. Moonshot AI’s Kimi K2, another open-source model, has further solidified China’s lead in the AI market. The US must recognize that its closed-model approach is at risk of being eclipsed by China’s open-source models.
China’s soft-power push in AI is focused on sharing and promoting its advanced technologies and infrastructure to countries across the developing world. This strategy has resonated with populations in developing countries, who are more optimistic about the economic and social benefits of AI than their counterparts in advanced countries. The US must capitalize on this optimism by prioritizing open models and promoting collaboration between model developers and research institutions.
The US government should recalibrate its AI export control strategy to align with the current direction of AI development. While export controls cannot prevent Chinese companies from producing advanced open models, controlling the semiconductors most suitable for inference could become a central point of a future US export control strategy. By promoting open-model development and more carefully tailored export restrictions, Washington can advance American soft power.
Ultimately, the US must address China’s growing soft-power lead quickly to maintain its global influence. If it fails to adapt, Chinese diffusion of cheap, powerful AI capabilities could prove too hard to displace.
Source: https://www.foreignaffairs.com/united-states/chinas-overlooked-ai-strategy