As the US ups the stakes on AI, can China win over the world with open-weight models?
China is intensifying its AI diplomacy with a push for open-weight models, challenging the US-led Pax Silica coalition. President Xi Jinping opened the World Artificial Intelligence Conference (WAIC) in Shanghai, pledging Beijing's commitment to open-source AI and establishing the World Artificial Intelligence Cooperation Organization (WAICO). This initiative aims to position AI as an "international public good" and counter US influence over technology governance. Beijing startup Moonshot AI unveiled its Kimi K3 open-weight model at the conference, rivaling top US systems.
Moonshot AI's Kimi K3 open-weight model, launched at WAIC, directly challenges leading US systems like Anthropic and OpenAI. This release, alongside China's broader open-source push, reflects Beijing's strategy to build a Global South-centered counterweight to US tech controls. Chinese state media frames US restrictions as an "AI Iron Curtain," advocating for an open-source, all-factor AI ecosystem.
The open-weight model strategy offers a distinct advantage for Asian markets outside the US sphere of influence. Countries seeking alternatives to proprietary US models will find China's approach appealing. This could accelerate AI adoption in Southeast Asia, where companies may prefer the flexibility and lower cost of open-weight systems, fostering local innovation without dependence on US tech giants.
The key thing to watch is the adoption rate of Chinese open-weight models in emerging Asian economies over the next 12 to 18 months. If WAICO successfully mobilizes significant international participation and investment, it will solidify China's position as a viable alternative for AI development. Failure to gain traction, however, would limit Beijing's diplomatic leverage against US tech policies.
Related reading
6 stories
Huawei plans Q1 2027 launch of new AI chip as it takes on Nvidia
We previously covered Huawei's aggressive AI chip development, a key part of China's tech strategy.
Mark Zuckerberg and Nvidia CEO weigh in on Anthropic AI proposal
Mark Zuckerberg and Nvidia's CEO weighed in on Anthropic's AI proposal, a direct competitor to China's open-weight models.
China has gained pace in the space race by recovering rockets – but can it relaunch one?
UN turns to Google to make its global data ready for AI agents
Rival AI agents, Instinct and Meta’s Muse, both add the ability to make calls
